Blue flames Orange flames

Just like this guy's, the color of my stove's flames were affected by the humidifier as well.

Why does this happen? Is it a good thing or a bad thing ?

  • $\begingroup$ Comments are not for extended discussion; this conversation has been moved to chat. Please note that any further comments which are not suggestions for improvement of the question, or requests to clarify it, are likely to be deleted. $\endgroup$ – David Z Nov 19 '18 at 21:27
  • 1
    $\begingroup$ Given the explanation in the accepted answer, it'd be neat if you could post a third picture taken when the humidifier is filled with distilled water, which should lack the salts found in drinking water. Grocery stores often sell distilled water in gallon-sized plastic jugs. $\endgroup$ – Nat Nov 20 '18 at 7:54
  • 1
    $\begingroup$ The pictures in the question are not by OP. He took them from the linked website. So no chance we'll get a third picture with distilled/deionized water. $\endgroup$ – user27542 Nov 20 '18 at 10:17
  • 1
    $\begingroup$ Can you confirm or clarify if the color of these flames appears to be the same as that caused by normal cooling of the flame? You could spray or drip water that is not from the the humidifier to see if the color is the same. Putting very cold water in a thin metal pot over a high flame should also cause water to condense on the pot and drip into the flames, causing an orange colored flame for a brief moment. It would be edifying to know if the orange colors are the same. Photos of both flames taken with the same camera might be interesting also (although not necessarily conclusive). $\endgroup$ – Todd Wilcox Nov 20 '18 at 21:32
  • 1
    $\begingroup$ Related. chemistry.stackexchange.com/questions/5000/butane-burning-color $\endgroup$ – aidan.plenert.macdonald Nov 21 '18 at 18:20

The explanation I furnish below will stand or fall on the outcome of an experiment I and others here have suggested which is also outlined in my response. I promise to edit or delete my answer per the recommendations of the moderators here if that experiment shows it to be wrong.

Humidifiers that operate on the "cold" principle- mixing tiny droplets of water thrown from the blades of a fan with a blast of air- produce a mist of water vapor-enriched air mixed with the partially-dried remains of water droplets that are enriched in salts by evaporative attrition.

Those salt-enriched specks, when drawn into a hot gas flame, then emit light at frequencies corresponding to the line spectra of the salt constituents. In the case of sodium chloride (the most common salt in tap water), the sodium produces a yellow-orange glow when it hits the flame.

This phenomenon forms the basis of a chemical analysis technique called flame spectroscopy, in which a platinum wire is dipped into a solution containing an unknown mixture of salts, and then stuck into a hot flame. The colors emitted as the salts in the solution are heated are then used to identify the chemical constituents of those salts.

(Since sodium is ubiquitous, and this test is so sensitive to it, the platinum wire must be dipped in hydrochloric acid, heated to redness, quenched in the acid again and reheated several times to rid it of sodium before running the test on the sample.)

This mechanism can be ruled in our out by observing the flame through a grating that separates out the primary sodium line and I invite anyone here who has a gas range (which I do not) and a grating (which I also do not, sorry) to perform the experiment and report back to us here.

Since any dust in the kitchen would likely have salt in it, if the humidifier fan is blowing dust into the flame it would make the flame yellow as well. This can be tested by running the humidifier without water in it.

  • 10
    $\begingroup$ How much salt is in your tap water? I find it unlikely that drinkable tap water would be brackish enough to show noticeable sodium colouring in a flame. $\endgroup$ – Anders Sandberg Nov 18 '18 at 12:18
  • 19
    $\begingroup$ @AndersSandberg: Then again, the sodium spectral line is very strong, and it takes very little salt to make it show up nice and bright. Back when I was doing flame tests in first-year inorganic chemistry lab, we'd spend quite a bit of time making sure to remove all sodium from the sample before even trying to test for any other elements, because if there was even the slightest trace of sodium left, it would overwhelm pretty much everything else. Oh, and don't touch the platinum wire with your fingers after cleaning it, because human sweat has lots of sodium in it. $\endgroup$ – Ilmari Karonen Nov 18 '18 at 12:30
  • 5
    $\begingroup$ This answer is almost certainly wrong. You can get exactly the same effect by holding the tip of a knife in the flame, or any other object. $\endgroup$ – Maury Markowitz Nov 18 '18 at 13:51
  • 17
    $\begingroup$ With all respect, the answer is almost certainty incorrect. The orange flame color is obtained in a gas flame in heating tubes in restaurants and public plazas by fuel-rich mixing. I took a pocket spectroscope to these, and their spectrum is indeed continuous. This is a thermal radiation from soot, not the Na line. I live near the ocean in a so salty air that all my tools corrode if I don't oil them for storage, as do guitar strings--and still my range flame is blue. I am very, very skeptical the OP's humidifier concentrates trace Na+ in air to express its telltale line suddenly so brightly! $\endgroup$ – kkm Nov 18 '18 at 23:10
  • 12
    $\begingroup$ @kkm : Then that simply gives two competing explanations, each of which would seem to produce the same result with a naive observing device like the eye and/or digital camera. The only way this could be analyzed more conclusively is, thus, to conduct an empirical investigation with an experiment to try and reproduce the effect as achieved with a humidifier nearby as in the OP,with spectrometer present to see if it is line emission or continuous (thermal) spectrum in that particular circumstance. It's entirely possible for the same (esp. naively) observed effect to have multiple etiologies. $\endgroup$ – The_Sympathizer Nov 19 '18 at 0:36

OK, this question appears to have generated some controversy. On the one hand is the answer by niels nielsen (currently accepted), which implies that the orange color is from sodium. On the other hand is the answer by StessenJ, which implies that the orange is normal black body radiation from the soot. Plus there are lots of commentators arguing about rightness or wrongness of the sodium answer.

The only good way to settle the matter is an experiment. I did it, with some modifications. First, instead of gas stove I used a jet lighter (ZL-3 ZENGAZ). Second, instead of humidifier I used a simple barber water spray. The third necessary component is a diffraction grating, a cheap one I had bought on AliExpress. I inserted it into colorless safety goggles to avoid necessity for a third hand.

When I lit the lighter I saw a set of images in the first diffraction order: violet, blue, green, yellow and some blurred dim red. So far consistent with the spectrum of blue flame given on Wikipedia. Then I sprayed water in the air, simultaneously moving the lighter trying to find the place where the flame will change color. As the flame got orange jets instead of initial blue, I noticed orange image of the flame appear between red and yellow images in the diffraction grating.

Below is a photo I could take with the grating attached to a photo camera's lens, having mounted the camera on a tripod and holding the lighter and spray in both hands while 10s exposure was in progress (sorry for bad quality). Notice the yellow/orange (colors are not calibrated) tall spike at the RHS: that is the part only present in the orange flame. (The jet indeed became visibly taller when it changed its color to orange.)

photo of the flame image

From this follows that the orange color indeed comes from sodium, otherwise the orange flame's image would be much wider and spread into multiple colors like the flame from a candle or a non-jet lighter.

The readers are welcome to replicate this experiment.


OK, I've managed to measure some spectra using my Amadeus spectrometer with custom driver. I used 15 s integration time with the flame about 3-5 cm from the SMA905 connector on the spectrometer body.

Below the two spectra are superimposed, with the blue curve corresponding to the blue flame, and the orange one corresponds to the flame with some orange. I've filtered the data with 5-point moving average before plotting. The spectrometer has lower sensitivity near UV and IR, so disregard the noise there.

(Click the image for a larger version.)

flame spectra

What's worth noting is that not only the sodium 590 nm line is present in the orange flame, but also two potassium lines – 766 nm and 770 nm.


Just tried the same with a humidifier instead of the spray. The result with filtered tap water is the same: orange flame with sodium peak. With distilled water, although the experiment with the spray still resulted in orange flame (basically the same as with tap water), with the humidifier I got no orange at all.

Anyway, in no one case was I able to make the lighter emit continuous spectrum. Whenever I got orange flame, it always appeared to be sodium D doublet, not continuous spectrum.

  • 24
    $\begingroup$ I suggest trying this experiment again, but using distilled instead of tap water. $\endgroup$ – David Hammen Nov 19 '18 at 20:35
  • 2
    $\begingroup$ Very nice! Though by this experiment we can't really tell if it's sodium specifically. It might also be calcium. It would have been great to do the experiment with a more accurate spectrometer. $\endgroup$ – jkej Nov 20 '18 at 9:28
  • 4
    $\begingroup$ @DavidHammen I've tried the experiment with distilled water (the bottle label reads "дистиллированная вода более 30%", whatever that "more than 30%" means...). The results are the same: I still do get the orange spike in the spectrum. Maybe the water is not pure enough, dunno... $\endgroup$ – Ruslan Nov 20 '18 at 18:29
  • 3
    $\begingroup$ @ToddWilcox I don't quite understand how you would get 590±2 nm line with incomplete combustion. What combustion product would emit this line? $\endgroup$ – Ruslan Nov 20 '18 at 21:54
  • 6
    $\begingroup$ @ToddWilcox It could certainly be improved, but this experiment clearly demonstrates that it is possible to induce sodium line emission in a gas flame to the extent that it appears orange by spraying the air with tap water. That goes a far way to settling the question in my book. Are you saying that something else than sodium might have caused the emission line that just happens to be at the right wavelength for sodium? Well, it certainly isn't black-body radiation from soot, and until someone comes up with a more plausible explanation, I think it's fair to assume that it's sodium. $\endgroup$ – jkej Nov 21 '18 at 14:53

The water cools the flame to the point where you get incomplete combustion, just like a candle. The yellow light is from glowing carbon, a.k.a. soot.

  • 17
    $\begingroup$ Soot gives a black-body spectrum, salt gives the Na atomic lines at 589 nm. So this would be easy to decide experimentally. Look at the flame through a grating. Or a CD disk. $\endgroup$ – Pieter Nov 18 '18 at 10:53
  • 6
    $\begingroup$ Also, sodium yellow is a fairly recognizable flame colour. The more orange colour in the photo, assuming camera colour correction have not completely changed it, looks much more like incomplete combustion than sodium to me. $\endgroup$ – Anders Sandberg Nov 18 '18 at 12:17
  • 5
    $\begingroup$ the original Golden Gate Bridge overhead lights were low-pressure sodium, which produced an intense orange color that is very similar to the orange tint in the OP's photo. modern high-pressure sodium lights are whiter in color because they have a thermal spectrum superimposed on the line spectrum. $\endgroup$ – niels nielsen Nov 18 '18 at 19:01
  • 1
    $\begingroup$ @AndersSandberg the camera often lies! (Though I agree with your conclusion) $\endgroup$ – Chris H Nov 19 '18 at 9:26
  • 4
    $\begingroup$ @nielsnielsen I'm an experimentalist (in fact a spectroscopist) and a large part of me hopes that no one has done the experiment byt he time I get home tonight. I lack a spectrometer at home but dont lack old CDs, cameras, or lenses, so could bodge something. $\endgroup$ – Chris H Nov 19 '18 at 10:10

The accepted answer is not correct.

I have a gas stove in the basement which I have to do periodic maintenance on. This requires you to remove a bunch of fake logs, which are made out of some very lightweight material, I think something similar to rockwool but more solid. After maintenance I turn it on to be sure it's still working, and noticed that if the "logs" are not in, the flame is pure blue. This piqued my interest, so I replaced the logs and noticed that the flame turns orange after a short period, which corresponds visibly to the "logs" beginning to glow red. For instance, here is the stove shortly after starting (as quickly as I could run from the thermostat to the stove) and then again about two minutes later:

enter image description here enter image description here

Not a huge difference, but you can see it. The flame along the front has no "log" over it so it remains blue. There used to be some rockwool insulation here but I removed it thinking it was left over from the installation.

It is not entirely clear how the "logs" do this, but it is clear this is purely due to the temperature of the flame. For further proof, I took these two photos of our cooktop:

enter image description here enter image description here

As you can see, simply inserting something cold into the flame causes it to turn orange. Now a huge effect here either, but that's because I was one-handing the photo and the knife isn't properly positioned. The humidifier does this by inserting a mist of water over the entire area.

  • 15
    $\begingroup$ "the knife will have sodium" - of come on, at the absolute best-case limit it will not be enough to be visible, and especially not enough to explain the fact that it will continue to be orange all day long if you care to hold it in that long. And what do you think is happening in the stove? Those "logs" are almost 30 years old and burn orange as long as the stove is running. The sodium explaination is completely implausible in both instances, or the hundreds of other objects you might try. Have you actually tried this, with anything? $\endgroup$ – Maury Markowitz Nov 18 '18 at 14:53
  • 10
    $\begingroup$ @MauryMarkowitz "especially not enough to explain the fact that it will continue to be orange all day long if you care to hold it in that long" .. if the temperature of the object really was the reason, shouldn't it have heated up enough for the flame to lose that orange colour by then? $\endgroup$ – muru Nov 19 '18 at 7:04
  • 6
    $\begingroup$ This might be correct; but the justification seems extremely weak to me. If the issue was with the temperature; why don't the logs start with an orange flame when the temperature distribution is greatest? Why does the knife keep the flame glowing red even when it's allowed to reach closer to equilibrium. I think this is closer than the accepted answer; but at the same time isn't really any more justified. $\endgroup$ – JMac Nov 19 '18 at 15:23
  • 8
    $\begingroup$ @MauryMarkowitz You have clearly never tried to stop experiments being upset by sodium flame contamination. First, sodium is everywhere (glass, for example) and a very, very small amount gives a visibly detectable flame colour. The logs are probably made from a sodium-containing ceramic. Don't speculate about things you have no experience of. Every glassblower knows you are wrong in practice. $\endgroup$ – matt_black Nov 19 '18 at 21:15
  • 6
    $\begingroup$ This answer doesn't really say anything as to why humidifiers cause the flame to change color, it just shows that you can also partially change the color of a flame by inserting something into it, but does not prove that the mechanism by which the knife/logs change the color is the same as the one from the humidifier. $\endgroup$ – Herohtar Nov 20 '18 at 18:54

Little bit of every thing here. But I would certainly not ignore CO. Remember gas comes out under pressure, and if any passes through the heat of the flame before being completely oxidized, you get CO. Presence of sodium could contribute, I would test with distilled water first. But water droplets would cool the flame faster by drawing heat to evaporate. Get a CO meter, and open the windows!


The humidifier increases the percentage of water vapour in the air, which decreases the percentage of oxygen. The deficit of oxygen makes the burning of the gas less efficient, with the orange flame that also indicates a higher amount of CO in the exhaust gas.

Make sure to increase ventilation when you see orange tips on the flames.

  • 20
    $\begingroup$ The humidifier surely isn't displacing any significant amount of oxygen. $\endgroup$ – David Richerby Nov 18 '18 at 12:24
  • 4
    $\begingroup$ If you're going to disagree with me, do it with metrics. $\endgroup$ – Douglas Held Nov 18 '18 at 16:45
  • 16
    $\begingroup$ That's why there are so many reports of humidifier deaths, right? $\endgroup$ – immibis Nov 19 '18 at 3:41
  • 8
    $\begingroup$ @DouglasHeld: There is 21 % of Oxygen in air. The saturation vapor pressure for humidity corresponds to approx 3 % of absolute concentration at room temp. (beyond that there will be "rain"). So there will not be a significant decrease of oxygen. It is not possible. Practically the humidifier will also not be able to come even close to the 3% (more likely between 1 - 2 %). $\endgroup$ – Andreas H. Nov 19 '18 at 16:32
  • 7
    $\begingroup$ @user71659 You can't compare pressurized pipework leaks with superheated vapour to a tabletop humidifying unit. Those are not humidifiers, and do not act like them. In that case, the pressure is built up far above atmospheric, so the vapour can take up a far higher percentage of the air which is kept at a lower pressure. You need situations like exploding heated and pressurized pipes in fairly enclosed spaces to have something like that happen. $\endgroup$ – JMac Nov 19 '18 at 21:05

protected by David Z Nov 19 '18 at 2:18

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?

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