Would 3/4" pex from a tankless water heater 25' up in the attic get hot water to the kitchen sink on the 1st floor faster or slower than 1/2" pex?

  • 3
    If you add insulation to the pipe it won't take so long if you draw water twice within, say, half an hour. Dec 5, 2023 at 17:02
  • 5
    @Ecnerwal, it's not likely the OP has a choice where the water heater needs to be located WRT the faucets. My own tankless is in my garage and I have no attic to put it in. My kitchen sink is probably also about 25' away, due to how the house was built in the 90's. Fortunately, that's the furthest faucet in my house, but "why would you..." is easily answerable with "because I didn't have a choice" or "because I put it where the old water heater was". Unfortunately, your question is fairly condescending. Dec 7, 2023 at 21:18
  • 3
    There's a question of English language here: "Faster" can equally mean both "sooner" and "at a higher rate" (and slower can mean the opposite in both cases). It seems reasonable to assume that 2sooner" is the desired interpretation, but really it should be clarified
    – Chris H
    Dec 8, 2023 at 10:24

3 Answers 3


Slower - 3/4" pipe has more internal volume, (approximately double) so when the water in the pipe has cooled down and you call for hot water, you have to flush out about twice as much water before the warm water gets to you.

  • 1/2" PEX ID 0.485"
  • 3/4" PEX ID 0.681"

Volume/100 ft

  • 1/2" 0.92 gallon
  • 3/4" 1.83 gallon

Specs source page no affiliation


  • 1/2" 0.23 gallon
  • 3/4" 0.46 gallon

With 2.2 GPM faucet,

  • 1/2" 6.3 seconds
  • 3/4" 12.5 seconds
  • 3
    The OP my be confusing the time it takes hot water to get to the delivery point with the flow rate (gal/min) that can be delivered at the use point. A 3/4" line would deliver a higher flow rate than a 1/2" line IF the water heater can supply that flow rate and if a restrictor at the faucet isn't limiting. Usually a 1/2" line is sufficent for a single use point. And the time delay is half of that for a 3/4" line. Dec 5, 2023 at 16:17
  • 10
    The restrictor on the endpoint is almost always the limiting factor, and the question was about time, not flow rate. But the flow rate will not likely change a whit, as the restrictor on the endpoint (or even just the orifice on the endpoint valve if the restrictor is removed) will rule.
    – Ecnerwal
    Dec 5, 2023 at 16:36
  • 7
    And since the question is about time to hot water, they will still get hot water twice as fast with 1/2" PEX.
    – Ecnerwal
    Dec 5, 2023 at 17:22
  • 5
    @JMac it is approx. an order of magnitude. 23 GPM @ 3/4" vs a 2.2 GPM faucet. resources.hy-techroof.com/blog/… Dec 6, 2023 at 22:29
  • 3
    @JMac It's you not understanding how plumbing actually works. And asserting things that not true as a result. Let's say 50 PSI supply pressure. 25 feet of pipe. 2.2GPM faucet. 1/2" pipe, 49.425 PSI at the input, 0 at the output. 3/4" pipe, 49.855PSI at the input, 0 at the output. No Practical Difference engineeringtoolbox.com/pe-pipe-pressure-loss-d_619.html
    – Ecnerwal
    Dec 6, 2023 at 23:07

Let's talk about extremes to discuss the meanings.

Let the thick tube be 1" and the thin 1/4".

If you consider "slowness" as ammount of cold water needed to be dumped before the hot appears, then thicker tube is slower. The thick one is having 4 times bigger diameter and 16 times bigger cross-section. In other words, every meter (inch, feet,...) of thick pipe holds 16 times more water tahn the thin one.

If you consider "fastness" as ammount of water that flows through the pipe per second (minute, hour,...) - flow rate - then the situation gets a bit tricky. The flow rate is driven by pressure differences and obstructed by pressure losses. The thicker the pipe is the lower pressure loss in the pipe. On the other hand, if you have 1/128" faucet, no matter how big the pipe is, the flowrate will be limitted. Also straight pipe will have higher flowrate than crazed labyrinth of pipes.
Rule of thumb here is "from thicker to thinner". Here the conduit on the heater sets the upper limit for your gauge.

The other point, not asked by OP but by Ecnerwal in comments is the ammount of wasted water by such a design.

What is actually wanted is shortest reasonable piping from the heater to the faucet. Short pipe reduces both waiting time for the hot water and water waste while waiting for it.


To answer the unasked, but very related question, of how to get hot water faster to your faucet(s), you could consider a recirculating pump. These are usually for very large houses and increase cost for heating water, but if you absolutely demand nearly instant hot water at the faucet, a recirculating pump might be your ticket. You'll also likely need to add plumbing in your house to make it work, so it's definitely not cheap and likely costs more than just the pump and labor to install the pump.

Take this example, for instance:

Or you could consider under sink tankless heaters. These can work to heat water until water from your main water heater arrives. These can still increase the cost of heating water, but don't require the kind of replumbing the recirculating pump does. And most of these are DIY ready, so you only need to buy the equipment, if you are handy with plumbing and electrical.

The trade off with this solution is that you do this for a single faucet at a time, instead of the whole house benefitting from it. But maybe only your kitchen needs it, so why spend more if you don't have to?

For example:

(Links are only for example. I'm not promoting brands or stores.)

And let's face it, the real/root question is how to get hot water to your faucets faster than it's currently happening. The asked question is an implementation question that can may not actually have an answer that addresses your issue.

  • This doesn't actually answer the question verbatim. It tries to help a situation which OP probably can only alleviate with pipe diameters.
    – Tim
    Dec 8, 2023 at 14:36
  • @Tim, there's nothing in the rules that state answers have to address the question verbatim. In fact, many of the best answers on all Stacks don't address the question verbatim. BTW, the answer by Crowley doesn't address the question verbatim. And I showed that the OPs issue could be solved by other means than pipe diameter, so no, your assertion there is also false. Dec 8, 2023 at 17:16

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.