Pyro gasoline

[Milutin]

In our refinery we buy pyro gasoline from petrochemical complex, we treat pyro gasoline on hydrotreating unit to remove olefins and after hydrotreating it is sent to Sulfolane plant, aromatics extraction.

Usually pyro gasoline color is between +30 and +20 saybolt, but we started to receive pyro gasoline with poorer color, even +4. In both cases color of pyro gasoline after hydrotreating is +30.

I’m afraid that we receive some impurities which can harm hydrotraeter catalyst. To try find cause of color changes we treat pyro gasoline with active charcoal, but no change in color. What is possible cause of pyro gasoline color change, or what is your suggestion for lab analysis to tray find reason for color degradation.

Regards,

Milutin

 

[25362]

As far as I can remember color is affected by the relative amount of conjugated polyolefins (gum precursors) and their molecular weight. UOP 326-06 gives the maleic anhydride value whichn tries to measure their quantities in mg/g.

May be ASTM D-381 and D-873 used to detect gum (existing and potential) can be of help.
In this respect you may visit:

http://gauss.ffii.org/PatentView/EP296646

http://www.freepatentsonline.com/4783416.html

 

[dcasto]

Also, you might check with others on this topic, but we'd see some off color caused by "Green Oil". Its a water soluible polymer cause by heating the gasolines in the presence of a caustic, like the caustics used to treat for small amounts of CO2 and S (all kinds of sulphur compunds)

 

[Milutin]

25362 and dcasto thank you for your answers.

This hydrotreating we use also for kero HDS and for pygas. Ten days ago as we introduced pygas for hydrotreating, dP started immediately to increase, from 3.6bar to 6.0bar for 5days . dP increase was constantly growing.

We also had this problem with color for about 3 days, after that color of feed was +20.

5th day in pygas operation we changed recirculating compressor for maintenance reason , in moment of change pressure had little disturbance, and dP sharply decreased to old value 3.6bar. After this dP started climbing again, and after 6 days it was 6.4bar.

Today we decided to switch to kero again and check what is reason for such continues dP increase.

In our hydrotreater we have two types of catalyst, first NiMo, and second CoMo.
I find out in literature if styrene concentration in pygas is high and if it is reach CoMo catalyst it starts to polymerize forming gums and rapidly increase dP in reactor.

Because of that I check styrene concentration in feed earlier in this year and now it is between 0.8 and 1 vol% in both cases. I thing this is not reason for high dP.

Now we have color of pygas +20 but dP is increasing anyway.

Regards,

Milutin

 

[0707]

I suppose you need to check for oxygen presence on the PG you ar receiving from the petrochemical complex. Maybe there is the key for your pygas color changes.

"The pyrolysis gasoline (PG) and its distillate fractions composition contain a rather high level of reactable unsaturated compounds (diene, styrene, indene, etc.) inclined to polymerisation during processing and storage. The most important factors for the oxidation, condensation and polymerisation processes are temperature, oxygen presence, contact with metals and light access. As a result of these reactions insoluble high-molecular compounds, known as gums, originate even at the stage of storage and the problem is transferred to PG distillation and PG hydrogenation sections"


good luck

Luis Marques

 

[25362]

Luis Marques is right.

Feed composition may change a bit: for equal end points and sulfur content, a bit more C[sub]5[/sub] corresponds with higher Bromine numbers and Maleic anhydride values.

Stage II is usually trickle phase at inlet and vapor phase at outlet. Do you have cool HP separator liquid injected between beds to control temperature rise ?

 

[Milutin]

As I mentioned post before, during pygas operation (gas compressor change) and in switch to kero our pressure drop sharply decreased to usual value. In my opinion if catalyst is fouled with gums or other material like particles or coke it is not possible to get lower dP without shutting unit down and catalyst regeneration.

I checked feed (pygas) density, distillation and composition on chromatography, we have slight changes, as expected but didn’t find anything unusual to earlier characteristics.

It is possible to have fouling but I don’t have experience of self healing catalyst. Also I cannot explain so rapid dP increase during pygas operation.


Regards,

Milutin

 

[Milutin]

I forget to mention, we have recycle hydrogen quench between catalyst beds.

 

[25362]

A changing liquid/vapor flow r[é]gime might affect the [Δ]P.

 

[epoisses]

The delta P increase you are observing is almost certainly due to gums formation. The fact that delta P dropped during compressor change is most probably due to throughput and/or regime change i.e. an artefact with regard to the problem you describe.

You have to act quickly now in order to limit the damage. I would advise to shut down the unit and carry out hot hydrogen stripping at 350 deg C minimum during, say, 24 hours. This should remove the polymers from the cat. During the stripping you can hopefully you measure a certain delta P across the reactor significant enough to see it move downward (at least I hope) to decide how long you want to continue stripping. Don't wait too long though before you carry out such a stripping, because the more polymers you accumulate, the more maldistribution there will be and the in-situ stripping will no longer be effective. The only remaining alternative would be ex-situ regeneration.

Meanwhile, don't forget to verify the quality of your C5- and C9+ fractions, at least if you take those out before the Sulfolane to stick them into mogas as certain refineries do. The gums may end up in people's car engines and make the valves stick, in which case your marketing organisation will have an unthankful issue to manage (I have actually seen this happening).

Now to prevent reoccurrence, I would first of all contact the chemplant if you have not done so, to find out where the deteriorated product quality comes from. It could be oxygen ingress e.g. to the intermediate tank if there is one, as 0707 wrote. Have the nitrogen blanketing checked, or install one if there is none (I know this is more easily written than done...).
There could also be operational changes on the chem plant fractionation train, e.g. entrainment of heavy, coloured components (gums precursors) might explain the phenomena. Work with the chemplant to find out what happened.

Lastly you might consider a toplayer on the cat that is specifically designed to catch dP promoting contaminants, but that would be rather end-of-pipe.

 

[epoisses]

just to be sure, are you seeing delta P increase across the 1st stage or the 2nd stage?

 

[Milutin]

Changing compressor caused disturbance, recycle gas dropped down for about 5000Nm3/h, reactor temperature profile went up, no change in unit throughput or some other major disturbance.
I don't know why hydrogen recycle flow went down after compressor change. (Actually we didn't change compressor because when we started spare compressor it has problem with electric motor and we started previous compressor again)

But dP in reactor sharply went down and started to climb, as I described earlier.

When kero is introduced pressure went down sharply for about 1.2bar, after that slowly for another 1bar in period of 36 hours, and after that another compressor change and dP went down to value even lower than usual, and stay on that value during kero mode.

We don’t have pressure measurement on each catalyst bed, only on rector inlet and outlet. My first thoughts was that top of first catalyst bed is plugged, to check this I started quench flow to reactor (we didn’t use quench during first 11 days of pygas mode), but nothing happened to dP, I expected if bypass some portion of hydrogen flow around first bed dP will go down.
I put some operating parameters for pygas and kero mode:

Temperatures(C):

Pygas Pygas (with quench) Kero
Rx inlet 271 274 295
Rx 1 bed 325 340 290
Rx 2 bed 333 348 295
Rx outlet 338 345 295

Hydrogen flow(Nm3/h):

Pygas Pygas (with quench) Kero
Recycle 19500 14500 13500
Quench 0 600 0


Operating pressure, CHPS 49.5barg

According to petrochemical company everything is ok with pygas and they don't know reasons for pygas color?!!
They store pygas in floating roof tank with membranes for sealing.

Regards,

Milutin

 

[epoisses]

OK, looking at these temperatures the dP problem is in the 2nd stage rather than in the 1st stage. (Re-reading your posts I see NiMo and CoMo catalysts which should have already told me it's the 2nd stage). So this is not the usual 1st stage plugged up scenario.

What is going on on the 1st stage (diolefins to olefins hydrogenation)? Is it part of your county or is it in the chemplant? I would tend to think that if there is a gums problem on the 2nd stage, there must be a HUGE gums problem on the 1st stage catalyst, and if there isn't, your problem is caused by something else. Or is there any intermediate tank where the quality deteriorates badly?

Let's talk about that compressor change then. Any data on what happened to the purge gas flow and purity? Could it be that the cat has been on far too low H2 PP during a certain time?

 

[Milutin]

First stage - selective hydrogenation is done in petrochemical company, after that routed in their tank, and pumped in refinery tank. Second stage hydrogenation is in refinery. Our hydrotreater has 2 catalyst beds NiMo and CoMo. Situation is similar like in article:

http://www.axens-group.com/pdf/prod...lective Hydrogenation Catalysts_imp_recto.pdf

During kero mode we have hydrogen purge flow about 500Nm3/h and recycle hydrogen has 82-85mol% H2, during pygas mode purge flow is 0Nm3/h, and recycle hydrogen has 71-73mol% H2. We started pygas mode again and decided to decrease pygas throughput to check hydrogen issue. We are not able to increase hydrogen make-up because MU compressor throughput limitation.

We are going to check with catalyst vendor for allowed gum content in pygas.

Regards,

Milutin

 

[Milutin]

Pygas analysis

We analyzed hydrotreater feed and products (pygas).

First set of data
Pygas - feed:
Existing gums: 0.9 mg/100ml
Potential gums: N/A
Induction period: 150 min.
Bromine number: 51.8

Hydrotrated Pygas:
Existing gums 0.5
Potential gums: N/A
Induction period: >360 min.
Bromine number: 0.13

Second set of data
Pygas - feed:
Existing gums: 0.8 mg/100ml
Potential gums: 303.4 mg/100ml
Induction period: >360 min.

In my opinion quality of feed pygas doesn’t look so bad.

Regards,

Milutin

 

[mfelzien]

Mulutin,

Interesting thread. To try and answer your initial question:

"What is possible cause of pyro gasoline color change, or what is your suggestion for lab analysis to tray find reason for color degradation.”

In the process of cutting C2 & C3s, I think the creation of gums is a process of Methyl acetylene or Ethyl acetylene oligomer formation. The literature states these are brown-red in colour.

As you suggested, contact your catalyst vendor and make sure this won't harm the catalyst. You might be looking for a new vendor if your supplier doesn't help more than they have.

Much luck

Mfelzien

 

[Milutin]

Most possible reason for color degradation is oxygen presence which initiates polymerization and gum formation. As described in earlier posts of 2536, 0707 and epoisses.
Evidence for this: when you complete Induction Period lab analysis, (exposing gasoline to heat in presence of oxygen) color of pygas sample is brown.

On address

http://www.criterioncatalysts.com/english/PDF/HYSTRIP.PDF

all interested in can find hydrogen stripping procedure. Also same site has very good articles about catalyst, hydrotraeting technology and hydrotreaters troubleshooting.

Regards,

Milutin

 

[Milutin]

Hi all
We have MAV of pygas, it is in range 1.9 to 2.5, does anybody have experience, is it that number too high for second stage hydrotreating?

Regards,

Milutin