Why does "dripping the water" work?

At our last house, the master bathroom was along the back wall, and the pipes for the shower were along that wall as well. They froze once - luckily not enough to burst the pipes. After that, we made a point of leaving them dripping if it was below 20 or so at night. When we moved out, there was a cold snap that first winter so we phoned the new owner and warned her of this.

One winter, though, even that might not have helped: the water froze outside the house. That day, apparently it was happening all over the county. We had to have the water company come out and thaw it at the meter in the ground in front of the house.

At a high enough flow rate, I’d agree. But a drip is not very much. It’s only a few watts of warming power for plausible rates and temperatures.

As I noted above, perhaps that is enough to prevent a partial freeze from turning into a total freeze. But that’s entirely speculative.

In fact it’s not clear under what conditions a freeze actually results in a burst pipe. Mama_Zappa says that she experienced a freeze without burst pipes. Does that mean it only froze in a short segment, which wasn’t enough to increase the pipe pressure enough to burst? Is there some “luck” factor involved?

I did find this company selling freeze protection valves. But the valves seem to go wide open while the water temperature is too low–that’s a flow thousands of times greater than a drip. Their protection is certainly entirely due to water supply temperature. A drip is probably a different story.

IME, it’s not a matter of leaving the tap barely open so it produces only a slow drip with gaps inbetween; but of leaving it open enough to produce a continuous stream, though a very small one, so as not to overstress the well or septic system.

I’m pretty sure the factor is that the water’s coming in generally at 50ºF or warmer. I very much doubt that “flowing water doesn’t freeze” because I see frozen waterfalls all the time. It’s possible that moving water doesn’t freeze quite as fast as still water, though.

I’ve also had freezes without the pipes bursting; more than once. I think it’s partly a matter of the material and condition of the pipes, and may also be affected by how hard the freeze is, and how sudden, and what the surrounding material is like (if the pipe’s coming through an also-frozen wall, for instance, there may be pressure from the surrounding material helping to brace the pipe.) But I don’t know for sure what the factor is; though I know for sure that it happens.

FWIW, the pipes I’ve had freeze and not break were all copper pipes; but I think I also had a copper pipe break, quite a few years ago.

I’ve had outdoor spigots freeze without breaking the pipes. I’ve had to warm them with a torch or heating pad to get the water flowing again. I assume in that case, the pressure of the new ice moves down the pipe towards the street rather than pressing against the sides of the pipe. If instead the ice was forming away from the spigot and moving towards it, the water between the ice and spigot wouldn’t have anywhere to go and the pressure of the ice could rupture the pipe.

That conflicts with what bobot said:

Perhaps the stream is required for deeper freezes. Or maybe the drips are sufficient but many homeowners turn it up to a stream since it probably doesn’t hurt (aside from the waste). I wouldn’t expect many homeowners to experimentally find the minimum flow by turning it down until their pipes burst, then backing off slightly.

I think even drips are probably enough flow, and the heat in them is a good enough explanation. A little drip irrigation thing that drips fairly slowly is still doing about a gallon an hour. A gallon an hour overnight is ~8 gallons. 8 gallons of ~45 degree water will melt quite a lot of ice!

Another way to think about it is how long the water in the pipe stays in the freezing region.

A gallon is 230 cubic inches. Water supply pipes are usually 3/4" wide, which means a cross section of .44 in^2. That means the water in your pipe is moving about 40 feet an hour. It’s going to depend on your pipe layout, but in many houses the section of the pipe exposed to freezing temperatures is short enough that the water is going to spend a lot less than an hour in freezing temps at that rate. Is less than an hour long enough to freeze a 3/4" tube of water solid? Not at 20 degrees it’s not. At 0 or -20 maybe.

ETA: fixed my math (maybe. Someone check me. Doesn’t feel right).

All true; and the last true in particular because conditions are likely to be a little different each time there’s a possibility of pipes freezing, so that determining the amount of flow that was or wasn’t barely enough on one night wouldn’t establish it for other nights, even those with similar weather reports. If the report says it’s going to be five below, it may actually be two below or eight below; and the wind direction and/or speed may vary; and the duration of the low temperature may vary.

That seems like an overestimate. An average water drop is 0.05 ml. Twice a second is 0.1 ml/s, or 360 ml/h. That’s about a tenth of a gallon per hour. A whole gallon per hour might not quite be a whole stream, but would be significantly more than a typical dripping faucet.

In the interest of SCIENCE, I set my kitchen faucet dripping into a 4c measuring cup and set a timer. The drips are coming about every 0.6 seconds. It took about 13 minutes to fill a cup, which is a flow rate of about 0.6 gallons per hour. There is splatter up to the very top of the measuring cup, so at least a very tiny amount of water is not being captured in the measurement. I expect a small enough amount that it’s not meaningful, but I noticed it so I’m mentioning it.

So, true wisdom of crowds, the actual rate is very close to the average of our two estimates. That still (assuming the rest of my math is right) means that the water in the pipe is moving ~24 feet an hour. That’s a lot!

In the interest of MORE SCIENCE, I repeated your experiment. At ~2 drops/s, I got 160 ml after 6 minutes, or 1.6 l/h, or 0.41 g/h. Less than what you got, but significantly more than my initial estimate! It seems that a “standard” water drop is much less than a real-life drop formed by gravity. For a 10 C difference, that’s almost 20 watts of heating power. So, maybe the heating theory is plausible after all, or at least contributes more significantly than I initially thought.

Let’s see. I use drip tape in vegetable crops. Assume 12" emitter spacing, 0.5 GPM per 100’ of line (not unusual) in which there’ll be 100 emitters, .005 gallons per emitter per minute; which is 0.3 gallons per hour. And each one does a pretty steady drip, IME. A gallon per hour is going to be a very fast drip.

I think (but am not sure) that it’s more than just the heating power. The moving water isn’t just like a coil around the pipe that’s heating the pipe/still water, it’s pushing the cold water through the pipe and out to somewhere warmer.

That’s effectively the same thing, though. It’s displacing liquid water at (a minimum of) 0 C for liquid water at 10 C. If there’s over X watts of outside cooling power, it can keep up and still freeze the pipe. Otherwise, the flow should dominate. And really, this is sort of best-case–if the flow is faster at the center than the boundaries, then the outer part could still freeze.

The one thing nobody is mentioning is: What is the temperature outside? Sure, I tiny drip might not prevent the pipes from freezing in the Antarctic, but if it’s only 22°F outside, it is probably enough.

Several people have mentioned this.

When I do the faucet drip thing (as I did last night), I noticed that when I first turn on the faucet, that first the water gurgles out for a couple of seconds, before producing the normal stream. Why does this happen?

Does it happen every time you turn on a faucet, or only when the weather’s really cold?

If the latter, your pipe may already be partly frozen when you first turn it on, disrupting the water flow; but as the warmer water flows through it, the ice melts and the stream goes back to normal.

The times I’ve had water do this it’s because the flow had previously been interrupted, though not necessarily by freezing – having had it shut off for any reason can do it; at least, if you’re describing what I think you are.

It only happens when I first turn a faucet on after a night of letting it drip. I don’t think the the pipes freeze, as I have a big house, so I do this at each sink (5 of them in the bathrooms and kitchen). And it never happens for the faucets that I didn’t let drip. For example, if it’s not going to be below 10 degrees F, I don’t do it with the basement faucets.

There can be fluctuations in pressure throughout the water system, especially if a fire hydrant is opened or large pumps are started and stopped, so that can suck a little bit of air back into your pipes from a slightly opened faucet. I suspect that’s what’s going on. Newer water systems have backflow preventers installed in customer service supplies to stop water flowing the wrong way and potentially contaminating the mains, but they’re by no means universal.

The simple answer is that a sufficient amount of flowing water above freezing temperature through a cold section of pipe adds enough heat in that area relative to the cooling rate outside the pipe to keep water in that section from freezing completely through, which would allow potential bursting via the expansion of freezing water in a closed system.

Flowing water does freeze; it freezes when sufficient energy/heat is removed, and can do so in a few forms not everyone has seen. Most people are familiar with surface ice forming on rivers when the atmosphere gets cold enough. You can also have conditions that allow frazil ice to form, which are patches of slush that form and keep moving with the flowing water. Anchor ice can form on the bottom of a stream, even when the surface is not frozen. An aerator keeps a small section of surface ice open by creating a slow steady flow of deep (relatively) warm water at 0-4C up to the surface and prevents or limits the cold -20C atmosphere from forming surface ice.

Once water starts freezing in a pipe or any other partially or fully enclosed area, it may or may not stick/expand/burst the confinement based on the shape of the vessel, how smooth the ice-wall interface is, how fast the ice forms, and how strong the confining material is. That’s why sometimes a cheap plastic outdoor spigot will burst when a very sudden cold snap comes through, while a strong metal well-insulated one can survive a mild winter with no issues. It doesn’t always happen at a certain temperature, and cracking the faucet may or may not work depending on variables.

The simplest experiment I know of that demonstrates this is a set of metal tubing in my yard. 3/4" pipe formed into an upward facing U-shape, open at the ends. During the summer it rains and the pipe partially fills with water. That same water froze during the winter. Despite being completely open (no-pressure) on both ends, a section of the pipe burst open when the water froze. The water froze in a way that created a plug which prevented it from expanding length-wise through the rest of the pipe.

Anyone who’s ever tried to just push a cylinder of ice out of a frozen a pipe knows how strong the friction is. Doesn’t matter if you leave a little crack in the faucet to reduce the pressure. Once ice forms across the full cross-sectional area of the pipe you’re pretty much screwed unless something warms it up before the material fails.

Dripping the water works by keeping enough of the pipe thawed in the cold exposed section via the heat of the incoming water to prevent it from freezing solid.