In Joseph Conrad's Youth: A Narrative, set in 1876, a 400-ton barque ship, the Judea, is sent to Java Head to transport 600 tons of coal. Although various delays occur during its journey, the coal it is carrying spontaneously combusts while off the coast of Western Australia. The crew attempt to smother the fire, unsuccessfully. Then attempt to flood the fire, but this is also unsuccessful. Finally, the gases in the hull explode and blow up the deck, and the crew hail a steamer to tow them to an intervening location. However, during their towing, the speed of the steamer's ship cause the fire currently present on the Judea to smoulder into flames, which the narrator describes thusly:

The speed of the towing had fanned the smoldering destruction. A blue gleam appeared forward, shining below the wreck of the deck. It wavered in patches, it seemed to stir and creep like the light of a glowworm

My questions:

  1. Given the amount of fuel and material of the ship available, is it possible that a blue fire of 2300-3000 degrees Celsius (from a quick Google search), could be produced within the hull of the Judea?

  2. Or is it the case that due to the nature of coal gas, a lower temperature than aforementioned need only be met for flames to appear blue, similar to how the flames of a gas stove appear blue?

Additional information, if helpful:

At its last port to Java Head, the Judea had been recalked and copper sheathed before embarking on the last stretch toward Java Head. The Judea explodes and is towed at night.

  • 2
    $\begingroup$ Copper could add blue/green color to the flame.. $\endgroup$ Commented Jul 13, 2020 at 20:29
  • 1
    $\begingroup$ Certainly case 2. A forge (coal with air actively blown in) barely reaches 1300°C. $\endgroup$
    – Karl
    Commented Jul 13, 2020 at 20:32
  • 1
    $\begingroup$ The temperature of the flame will not be the primary cause of its colour. Many gases will burn with a blue-ish flame under the right conditions (methane, hydrogen and carbon monoxide, for example, all potentially associated with decomposing coal). $\endgroup$
    – matt_black
    Commented Jul 14, 2020 at 15:21
  • 1
    $\begingroup$ I will upvote your question as the extent of background and commentary will likely reveal and support the right answer. The actual underlying chemistry, whoever presents it, will hopefully enlighten and possibly serve as valuable lesson for others. Thank you. $\endgroup$
    – AJKOER
    Commented Jul 14, 2020 at 18:01

2 Answers 2


In a Bunsen burner, the gas (methane or propane) may burn under two practical regimes : reductive or oxidizing conditions. In the reductive conditions, the flame is yellow and unstable, because the amount of oxygen is low with respect to the stoichiometric ratio. In the oxidative conditions, the flame is noisy, strong and stable, and it is due to an excess of oxygen in the mixed gases before the combustion. The oxidative flame is nearly colorless, but it has a characteristic blue cone at its base. This blue cone has a temperature of about $900°$C, and its spectrum has been analyzed by Hertzfeld in $1935$ : it shows a line spectrum typical of the diatomic "molecule" $\ce{C_2}$, with plenty of vibrational and even rotational sublevels. Maybe that was the origin of the blue flame described by Joseph Conrad. Why not ?

  • 4
    $\begingroup$ See also 'coal gas' - en.wikipedia.org/wiki/Coal_gas - produced by heating coal. $\endgroup$
    – Jon Custer
    Commented Jul 13, 2020 at 19:41
  • $\begingroup$ Maurice: I also enjoy reading your answer. Unfortunately, my expanded answer (see comments after Edit) indicate the failure of your response on addressing the likely chemistry, and especially on explaining the blue reported flame coloration. Further, my rendition does account for the contributing factor to the ship's disaster, an important point for future ship fires, relating to the interaction of hot iron and steam (from the unsuccessfully attempt at flooding the fire ). $\endgroup$
    – AJKOER
    Commented Jul 14, 2020 at 17:10
  • 3
    $\begingroup$ @AJKOER Is it so necessary to push your answer to this extant? feels excessive. $\endgroup$ Commented Jul 14, 2020 at 17:59
  • 2
    $\begingroup$ @ Ajkoer and Safdar. Two remarks. 1. 900°C in a flame is not considered as a high temperature. 2. Blue flames can be obtained by burning practically any coal gas, or any mixture of hydrocarbons with or without CO. My point is that this color may be due to the presence of C2 in the flame. That's all. $\endgroup$
    – Maurice
    Commented Jul 14, 2020 at 19:33
  • 1
    $\begingroup$ @AJKOER I am not talking about the correctness or validity of both answers. I am in no position to judge this. Rather the approach that you have taken to declaim Maurice's answer just felt like it was over the top and unnecessary. $\endgroup$ Commented Jul 15, 2020 at 4:10

The observed blue flame and explosion (the latter following per the quote, "Finally, the gases in the hull explode and blow up the deck") are likely the products of a hydrogen and air explosion.

There are two potential sources for the hydrogen gas here. First, the coal fire heats the mass of coal releasing a gas mix, to quote Wikipedia on Coal Gas on the particulars:

Coal gas is a flammable gaseous fuel made from coal and supplied to the user via a piped distribution system. It is produced when coal is heated strongly in the absence of air.

And further:

Coal gas contains a variety of calorific gases including hydrogen, carbon monoxide, methane, ethylene and volatile hydrocarbons together with small quantities of non-calorific gases such as carbon dioxide and nitrogen.

The second possible source of $\ce{H2}$ is from the failed attempt to flood the fire producing a large quantity of steam in the presence of hot iron, to quote an educational source:

If you pass steam over hot iron the steam reacts with the iron to produce a black, magnetic oxide of iron called triiron tetroxide, $\ce{(Fe_3O_4)}$.

$\ce{3 Fe(s) + 4H2O (g) -> Fe3O4(s) + 4 H2 (g)(1)}$

The hydrogen produced in the reaction is swept away by the stream of steam.

So, the hydrogen is formed and is trapped in pockets in the hull where it may eventually mix with oxygen from air and finally, a hot splitter causes an explosive ignition forming a striking blue flame.

[EDIT] On the explanation of the 'reported' (albeit, not always actual) color of a hydrogen flame, I quote a source:

There are also several text-books on chemistry which assert that hydrogen burns with a characteristic faint blue flame. It is easy to prove, however, that the flame of pure hydrogen has no blue tinge whatever. The blueness so frequently associated with the flame of hydrogen is really due to the presence of sulphur as is shown in a little paper I published in the Philosophical Magazine for November 1865.*

Now, in the current context, in the presence of burning coal, likely containing a significant sulfur impurity, a pronounced more unusual blue appearance apparently was reported, which would be in agreement with the hydrogen gas impurity argument, as is, the mention of a possible very high-temperature flame.

Others believing that this was a more normal occurrence (associated with, for example, a flame "with characteristic blue cone at its base") are likely incorrect. Further, not accounting for the chemistry as to how such a quantity of hydrogen gas was likely amassed, and its potential implications for future ship's fire, is a significant, perhaps even egregious, omission. More profoundly, it lies at the very explanation for the occurrence of the reported explosion and any resulting loss of human lives.

  • $\begingroup$ Why should the flame of H2 be blue ? $\endgroup$
    – Maurice
    Commented Jul 13, 2020 at 21:13
  • $\begingroup$ Actually, sorry you brought up this point, as it exposes unfavorably your otherwise well-presented answer above. $\endgroup$
    – AJKOER
    Commented Jul 14, 2020 at 16:58
  • $\begingroup$ (+1) I will upvote immediately after posting this comment: your answer is plausible to me. Funny thing about the way the stack exchanges work: in several cases, I have posted detailed answers that I consider to be clear and demonstrably correct, yet they get minimal upvotes. Then other answers, that are clearly incorrect, get numerous upvotes and get accepted by the OP. Frustrating, but not much to be done about it: it is a voting model, for better or worse, and we volunteer our efforts. $\endgroup$
    – Ed V
    Commented Jul 14, 2020 at 22:30
  • $\begingroup$ I also upvoted. But beside H2 or H, answer such as that of Maurice and point 2 in a comment were what I thought at first. CO burns blu and there should be sufficient oxygen at see. Explosion can originate from every flammable gas pocket and even fine solid particles previously formed in the fire. Personally I upvoted both answers and that comment. $\endgroup$
    – Alchimista
    Commented Jul 15, 2020 at 10:17
  • $\begingroup$ I disagree on the idea that it was a hydrogen explosion. High grade coal would be prone to methane release, which can build up in the hold, one reason why modern regulations require online gas measurement for shipping coal. For those of you who have seen old style refinery flue gas burning at night, if it's "burning blue" that means methane and/or other light gasses are being fully oxidised, so Maurice has my vote on the blue flame. Pockets of trapped methane will also explode when the concentrations are right (~7-14% in air) when it comes into contact with open flame - ask any coal miner. $\endgroup$
    – Gwyn
    Commented Jul 22, 2020 at 22:19

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.