High Inorganic Chloride salt n LVGO 1

[DDSpetro]

Hello…Experts !! Happy New Year to all of you !!

LVGO –Light Vaccum gas oil is one of the product from Vacuum crude distillation column , high Total chloride especially inorganic chloride is reported , while checking at overhead receiver boot water chloride it remains only 1 or 2 ppm , while LVGO chloride data shows total chloride 15 ppm ,20 ppm some times 23 ppm , it varies from 5 ppm to 25 ppm ; out of which organic chloride is very less say , e.g. out of 20 ppm , 16 ppm is inorganic and rest is organic , desalter efficiency remains ~ 98 % ,LVGO draw temperature is 130 to 140 C; LVGO Return temp. is ~ 40 C ; NH4Cl Desublimation temp. delta online monitoring is there but it shows more than 20 C delta difference, lab analysis shows no presence of Fe,Zn,Al metals & absence of amine,ammonia chloride or not even Na,Ca ; It sounds to be HCl.Caustic dosing in upstream crude unit is ~ 200 to 300 lph.One thing which was observed that , whenever NH3 demands goes up in VDU Column ovhd (Dosing as per Overhead reciver boot water pH), chloride in LVGO goes down ;

What are the potential cause of this chloride ? It is very critical point of concern because this LVGO goes to VGO hydrotreater and it can cause sever corrosion in the downstream.


Thanks in advance to all of you.

 

[DDSpetro]

There is good amount of correlation has been seen with LVGO chloride vz. LVGO PA Flow as compare to return temp. which remains with in 2 C variation.BY looking at graph, is it possible to prove HCl condensation theory, mean while LVGO sample is again given for Cr,FeCl3,Ni,V metal analysis along with moisture analysis in the sample because HCl only can be present if moisture is present.

give your expert comment pl.

http://files.engineering.com/getfil...fc24c45a8d4a&file=HCl_Condensation_theory.ppt

 

[25362]

The presence of chlorides in LVGO, and transfer line corrosion, might indeed be independent/unrelated events.

Magnesium chloride, for example, if not removed by the crude unit's desalter, releases HCl by thermal decomposition mainly in the atmospheric topping column; one of the results may be the formation of ferric chloride.

I don't have the vapor pressures of anhydrous HCl, but since its critical temperature is 51.5^oC, it is reasonable to assume that, if formed by salt decomposition, it stays as a vapor at the temperatures encountered in the distillation columns of a crude unit.

In regard to possible naphthenic acid corrosion, one ROT would be to keep the crude oil blend at an acid number [≤] 0.5 mg KOH/g; therefore, crude oil blends with greater values might result in corrosion.

 

[DDSpetro]

Desalter outlet salt ptb Vz. LVGO Chloride - No correlation found.

http://files.engineering.com/getfil...le=LVGO_Chloride_Vz._Desalter_Outlet_salt.ppt

 

[DDSpetro]

mgcl2 is separately not monitored in desalter outlet ,0.6 to 0.8 PTB DESALTED SALT reported as NACL,hardly we can assume 30% mgcl2 again correlation is poor with desalted crude.

 

[25362]

Curiously the left half of the plotted points seem to correlate quite well.

 

[DDSpetro]

Yes, 25362, it was correct …how do we technically justify ?

It was observed in laboratery that in LVGO sample, chloride gets reduced by 60 % by water wash at room temperature.Static mixture followed by coalscer will be good option ?


25362- can you pl. elaborate about HCl will remain in vapor phase as you mentioned in previous post ..so, Low LVGO PA flow or High LVGO Return temp. will not be the solution is it like that ?

earlier first time chloride analysis in LVGO no, presence of Fe found , however,second sample is given for it.

 

[25362]

A water wash is what actually an electrical crude oil desalter does. However, efficient mass transfer, and addition of demulsifiers, do not manage to reduce the salt content by 100%; that's the reason why refineries install two desalters in series.

See thread124-136505 for tips.

It is apparent that HCl formed by hydrolysis and by thermal decomposition of salts at the (P,T) conditions prevailing in the topping and vacuum towers should be in the vapor state.

 

[DDSpetro]

So, Chance of corrosive product FeCl3 will be higher rather than HCl,it looks like but why again i am wondering why PA flow should correlate with LVGO chloride salts as attached in the slide.

In fact one datapoint/outlier was removed from Desalter outlet salt vz. LVGO Chloride & very good positive corrleation has seen between one of desalter train outlet salt & LVGO chloride as attached in the slide ; while other train outlet salt was not correlating with LVGO chloride ,other parameters are studied and attached for your expert comments if any one contributing like HVGO CCR,95 %,Caustic injection,VDU Column flash zone temp. in crude unit as some good correlation has been seen with chloride salt as attached.

http://files.engineering.com/getfil...-95e09a485a02&file=LVGO_Chloride_analysis.ppt

Once again thanks 25362 for expert tips.

 

[25362]

Thanks for the undeserved star. One could venture a logic hypothesis to go this way: the PA flow rate affects the condensation of volatile salts in the main tower. More PA more salts condensed.

 

[DDSpetro]

Hello,25362

Lab analysis received in terms of Moisture / Metal presence in LVGO ; Moisture to the tune of 1821 ppm @ 90 C is found . But Fe metals is <1 ppm while Na is 1 to 1.4 ppm observed as attached in the file.

It is really struggle to check the effectivness of Caustic injection ( 6.6 % solution) in desalter downstream

Pl. for your expert comments.

 

[DDSpetro]

http://files.engineering.com/getfil...7f05a&file=Metals_&_Moisture_in_chlorides.xls

 

[25362]

Water content seems to be dissolved water from stripping steam. Are U using a stripper for the LVGO? Please note that Na may also come with stripping steam.

 

[DDSpetro]

lvgo section does not have stripper,vaccum column is using around 13 tph lp superheated steam 190 c and in vdu furnace in each pass 750 kg/hr turbulizing steam.generally chloride and Na is not monitored ,however i wil look for lab analysis.

 

[DDSpetro]

One intresting observation is that LVGO chloride is very well correlating with upstream crude column overhead boot water chloride as attached.

http://files.engineering.com/getfil...loride_vz._Crude_Col._boot_water_chloride.ppt

 

[25362]

Although one should beware of falling prey to the common logical fallacy "post hoc ergo propter hoc", these results seem to support the hypothesis that both have the same source.

 

[DDSpetro]

crude unit return condensate conductivity remains 2.8 MICRO OHM,So, possibility of chloride,Na can be eliminated.further it has been seen that incoming 45 % Caustic has only 50 ppm NaCl,further it gets diluted up to 6.6% with condensate and further in plant with non phenolic striped water but 20 to 25 ppm in lvgo is higher , desalter wash water doesnt hv mg & ca hardness but outlet brine is hving mg ,ca hardness ,so,desalter removes ca & mg salt. how to prove to increase lvgo pa return temp as it involves entire column upset.

 

[786392]

Dear kindly Pardon me is the Conductivity measure micro mho's & was it a typo? or not!

Best Regards
Qalander(Chem)

 

[25362]

DDSpetro, your assumption of condensing HCl is not to be discarded prima facie.

The azeotrope HCl-water (normal BP: 108.6^oC, 20.2% w/w acid) due to the acid differing "activities" at low concentrations and varying T,P conditions may explain the difference of Cl- contents between LVGO and the overhead boot-water.

I suggest you look for more published information on the HCl-water azeotrope at reduced pressures.

 

[DDSpetro]

u r right 786392, 25362- ok ,meanwhile,i hv seen caustic injection quill is of alloy steel in the direction of crude flow,270 to 350 liter/hr diluted caustic ~ 3 % from 6.6 % ,no choking in nozzle ,(location in d/s of desalter ) & stoichiometry wise ok to nulify it as cause; high acid treatment chemical has chloride but 50 ppm in 8kg/kt of crude ,very negligible in terms of 25 ppm chloride in 100 mt/hr lvgo draw,it seems that crude is definately playing a role but i wil see HCL/WATER AZEO as suggested and to check caustic injection effectiveness. thanks 25362 for ur prompt , expert & memorable help.i wil revert soon.

 

[25362]

It is not necessarily an azeo composition, but info about the vapor/liquid composition at low pressures on the water- rich side of the HCl/water mixture may be of help.

Try to find the expected liquid concentration of the HCl-water mixture at T,P conditions, taking into account that the Cl concentration (if it is indeed all HCl) in the LVGO, is about 1-1.3% of the water content.

Perry VI gives a partial pressure for HCl at 40 ^oC of 0,00047 mm Hg for a 2% solution. Regretfully it doesn't show values for lower concentrations. The water-HCl mix is dissolved in the LVGO, thereby, IMO, its mol fraction in the liquid should be taken into account.