Heat exchanger outlet temp does not make sense (Sulfur recovery unit)

[Bain89]

G'day guys

We have a tube heat exchanger which cools process gas from the Claus sulfur recovery unit (inlet 340 deg C) to about 140 deg C (outlet). The process gas contains H2S, CO2. The waste heat is used to generate steam. The boiler feed water is 100 C, steam leaves at about 140 C.

We noticed the outlet gas temperature was 70 C for 1 month! Which didn't make sense, as the lowest temperature medium is 100 C. And alarm bells rang in my head for condensation in the exchanger!

Opened up the heat exchanger, sure enough bottom was full of water, leaks found in bottom tubes. As the boiler water (shell side) is at 3 bar and process gas is 0.1 bar (tube side), a leak would introduce water into the process gas. This no doubt caused sulfuric acid to form, and there was a sludge of FeS, pyrite, and similar on the bottom of the heat exchanger that was covered in water.

Tubes plugged and all is fine now, exit temp back up to 140 C. Turns out during start up they did not purge with N2, so H2SO4 was forming which probably caused the original leak. Then once water came in, it went from bad to worse.

Can anyone provide an idea how the gas is able to leave at 70 C, when all fluid flowing in are over 100 C? This has stumped us.

Many thanks,
Rob

Rob Muggleton
Corrosion, Oil & Gas

 

[georgeverghese]

Based on what you've told us so far, you've got me stumped too. Though I cant get my hands on heats of solution for H2S and SO2 ( nothing on the net or in Perry), I'd presume heat is evolved during dissolution of these soluble components when they leak into the tubeside. Have you spoken to the plant operators running this Claus unit ? - they often know the reason why "things go bump in the middle of the night", and they'd be reluctant to tell you at the outset, unless you're the Plant Operations Manager or similar.

By the way , the dissolution of H2S in water produces the hydrosulfide ion HS-, while that for SO2 produces the hydrosulfite ion HSO3-. In more basic solutions, you'd get the sulfide and sulfite ion respectively. But to produce the sulfate ion (which is the product of ionisation of H2SO4 in water), sulfur trioxide gas is required, and this does not form in significant quantities in the upstream Claus reaction as far as I know.

 

[Bain89]

Thanks George - To add to the timeline, they were doing a start up, and couldn't get the outlet temp over 70 C, and ran like this for a month. They did not purge with N2 during start up.

I am more of a metallurgist than a chemist, however, during the start up they progressively warm up the unit to avoid damaging the refractories.
This is done also with adding high pressure steam, and they are burning with oxygen in excess of the 1/3 ratio, which would form sulfuric acid? This is perhaps a minor detail, either way there was an acid that ate through the pipe wall and let water in.

Thanks for the reply!

Rob Muggleton
Corrosion, Oil & Gas

 

[georgeverghese]

Texts on industrial chemical process kinetics tell us that a catalyst V2O5 ( vanadium pentoxide) is required to increase the speed of conversion of SO2 to SO3, without which reaction rates are very low. This reaction also would require high pressures, based on the vant Hoff equation / Le Chatelier principle, to get decent conversion rates.
Would suspect sulfurous acid (which is the precursor to the HSO3- ion ) and sulfided water are both corrosive enough to attack most low alloy steels to cause tube leaks.

Could the corrosive conditions on the tubeside also have chewed up the thermowell and the temp sensor ?

 

[Bain89]

Damage to the sensor was also our first guess, however, there was nothing wrong with the instrumentation...

We were considering that a small leak would have let water into the process stream, making the air saturated.
The gas outlet pipe diameter is 16", vertical, and not insulated.
We theorized there could be condensation on the pipe wall due to the saturation with water, which would run back down the pipe towards the exchanger gas outlet.

This condensate running back down the pipe wall may have cooled the temperature sensor. However, upon checking, the temperature sensor is right in the centre of the pipe, not at the pipe wall, so condensation would not be running over the temperature sensor making it read a lower value...

But yes you are right, either way it is a corrosive solution for steel. The outlet temp remains a mystery though.

Rob Muggleton
Corrosion, Oil & Gas

 

[georgeverghese]

If the exit line from this HX is not lagged, yes, I can see water condensate cooling down here at this TT sensor, if it were to be a low point in the piping where a water pool could build up.