Temperature Inversion 2

[Petro0707]

Hi All...

Recently we are facing problem of Tempearature inversion between naphtha draw off & crude fractionator column top temperature.

In normal condition – Naphtha draw off temperature remains around 130 °C & Crude column top temperature remains around 111 °C ( Delta between this two temp. ~ 19 C )

But abnormally after the occurrence of this event as shown in graph, naphtha draw temperature is remaining around 118 C & column top temperature is remaining around 114 C ( Delta ~ 5-6 C) along with sudden increase in Naphtha section Pressure Drop (PDI ) also around 1000-1100 mmwater column ( as shown in the graph)

Pl. see attached graph image by clicking, which indicates trends as specified below:

1.Green colour – Crude column overhead temperature
2.Red colour – Naphtha section pressure drop
3.Pink colour – Naphtha draw temperature

In the trend analysis, you can see sudden dip in Naphtha draw temperature on 16th September.

Naphtha PA return temp. has fallen down by 5 C i.e. from 100 C to 95 C . There is no any significant impact on Naphtha distillation & End point.

I have gone through the thread124-178998.

We have got total 3 trays in Naphtha section which is specifically designed HI-FI trays of MOC Monnel.I simulated the Model but it shows that if efficiency of Trays got reduced in naphtha section then in that case top as well as draw both temp. should fall down but here top temp. is higher and draw temp. has fallen down so, is it related to Tray damage or Pump around reflux distribution nozzle damage..Still not sure.There is reduction of Naphtha PA Duty but there is no any major changes in other PA duties of HK, Diesel, HAGO. Please note that we don’t use cold reflux for the column since very long time.

Request for your help please on this…( Regret for long description)

 

[25362]

From the net:

...In the case of M-400, a chromium oxide layer is not available to restrict the transport of nickel to the solution. As a result, a thick layer of sulphide crystals grows on the surface of the M-400 alloy even when subjected to a mild acid pH excursion. Even trace concentrations of sulphide/sulphate under normal pH control are shown to react with alloys whose oxide films appear permeable to nickel transport.

 

[Milutin]

Petro,

Increased dP in naphtha section and lower top P/A temperature are caused by tray damage, tray damage is caused by corrosion and maybe with some other causes.

We have similar arrangement on our CDU regarding top P/A - cold reflux. When we want to line up cold reflux first we drain water from low point on that line.
Water is for sure accumulated on that line if it is not in operations for months.
We learn this from experience, in past, during one startup of that line water was in such quantity that when it reached column it immediately boiled (better word is exploded) and PSV on column top opened to relive pressure.

Such event can also damage trays, if they cannot resist downward acting force.
So it would be interesting to check if somebody try to line up cold reflux in date when you noticed increased dP.
You can check column pressure and cold reflux flow.

Also very useful for your analysis would be detail description of first four trays inspection during column inspection. Pictures if exist can be great help.

You should pay attention on: quantity of deposits and their location (deck or downcomer), missing valves, corroded tray deck segments (location on tray), missing or removed tray segments from their original position, are tray segments bent, if yes position upward on downward, are valves stick by corrosion, are tray segments firmly attached to deck....
Such analysis can help you to decide what really caused problems in column top.

Sudden increase in dP in your case can suggest that corrosion is not only reason for tray problems.

Regards,

Milutin

 

[Petro0707]

We have found lot of deposits in down-comer side , there was no any missing or removed tray segments from their original positions could not found,lot of corrosions found on top 1st tray and as you can see in the attached photo graphs , plugging of sieve holes, corrosion layer on Monel shell , corrosion was severe near or above TSR, tray segments are fasten / bolted….

Please find the deposits analysis of that was found in downcomer :

Appearance – Black amorphorous
Magnetic response – slightly
Toluene soluble – 13.77
Loss of ignition – 41.87
Ash : 58.13
Carbon – 4.9 ( It is really miserable from where Carbon is coming )
Sulfur ( as sulfide ) – 11.3
pH of 2 % solution – 4.5
So4- - 0.09
NH3 – 0.01

Metals – Fe – 7 %
Ni – 31.8
Cu – 13.03
Mn – 0.30
Cr – 0.02
Al – 0.01
Si – Not detectable
Zn – 0.19
CO – 0.02

 

[Petro0707]

Hi...25362 & Milutin,

We could find some potential causes as below...

On top trays if top pumparound flow gets doubled than design...is it having some significance in terms of increase in top surface velocity on 1st tray which can make paper thin 1st tray over a period of time or free water that exists on top section of column and that remains circulating through p/a loop and can cause corrosion or errosion..., One thing which was noticed that .. total reflux flow ( p/a flow+hot bypass flow) could be controlled during initial rise of delta p across naptha section and pdi was coming down and then again it was increasing..but yes...during heavy sudden peak of 2000 mmwc --total reflux flow was very high....Is cold reflux really needed as we don't operate since long time...instead of high total p/a + hotbypass flow ..Is any purge/slip stream really required from d/s of naptha p/a pumps to purge deposits/salts etc.from the loop...we keep 10 -12 C higher top overhead temp. than water dew point temperature...but what can be the actual dew point with presence of NH4Cl ,So3- etc..we could not find any PSV poping during this phenomenon..

 

[Petro0707]

http://files.engineering.com/getfil...ff2d-4eca-9f31-9361a910b396&file=DSC01950.JPG

http://files.engineering.com/getfile.aspx?folder=8a9696f5-696e-4231-8311-346b884550a0&file=6.zip

 

[Petro0707]

http://files.engineering.com/getfile.aspx?folder=fa7ca5ae-341f-4040-95a2-6ddcabbd0292&file=T111.ppt

http://files.engineering.com/getfile.aspx?folder=f63b096e-5f4c-4b6d-99ba-173ca177dd15&file=T222.ppt

http://files.engineering.com/getfile.aspx?folder=68f2af68-f0ed-4487-97ca-16e4e10505fa&file=T333.ppt

 

[25362]

As for the subject of external reflux, it is the internal reflux that determines the degree of fractionation (ASTM gap) between adjacent fractions. Although the top 4 trays are for heat exchange, they are, anyway, considered equivalent to one tray for fractionation purposes.

I suggest, if possible, to read the pertinent chapter in D.S.Jones' Elements of Petroleum Processing (WILEY)

 

[Milutin]

Hi Petro,

From pictures you posted it is clear why pressure drop increased and naphtha tray collector temperature went down.
Sieve tray holes are plugged or partially plugged by corrosion deposits, that caused increased pressure drop in that tray section. Big holes or missing tray segments below top P/A distributor caused that one part of top P/A flow bay-passed tray deck and reached collector tray without proper contact with hot vapors. Corrosion and erosion caused by liquid stream from distributor are responsible for such big holes.

I don't thing that tray segments are so thinned in short period of time, for how long these trays are in operation? In addition, how much time passed from your last cleaning of these trays?

However, main question is what caused such corrosion on top trays? Most probable answer is free water on that tray. As you mentioned water dew point is 95degC, top P/A return temp. is 95-100degC. This can cause a lot of water to condense and absorb HCl, which is reason for corrosion. As you wrote you find chlorides in deposit analysis, (I am not sure how much).
To prevent, or at least decrease local water condensation you should increase top P/A return temperature, there are several ways:
- Increase top P/A flow, increase top P/A temperature (same duty)
- Increase other pumparounds, this will decrease top P/A duty, what will allow you to keep same top P/A flow and increase top P/A return temperature.


Regards,

Milutin

 

[Petro0707]

Dear Milutin & 25362,

Actually, within two years, trays are corroded.Because two years back as per the inspection report , everything was alright ,but one thing i would like to mention here that , as per the original design there was one stream from naphtha p/a pump discharge to tray down below ,Naptha Internal Reflux (IR) & we were having only two pumps used as p/a & top trays were valve trays ,but for some other reasons like Naptha End point,to minimize cost etc. we have modified & Naptha IR from p/a discharge to down below tray was removed , valve trays were replaced by Hi-Fi sieve trays,instead of original two pumps , we have added two extra pumps and now we are realizing that by putting extra p/a flows to top tray and that too..by looking at design of top distributor of p/a return to top tray it looks like that there is no equal distributions on top tray..may be possibility is there that at some locations instead of equal shower like distributions ,it may be impingment or localized distributions..

but as i mentioned earlier , we have added ester based corrosion inhibitor for two months (before 5 months back) in naptha p/a pump suction to minimize corrosion in naptha p/a cirucit , can this ester based CI cause glooey,sticky,slusshy deposits on tray..(as we observed)

we are still analyzing what ionic model/corrosion reaction is taking place inside top section ?

WIth in two months--fouling of naptha p/a exchanger & lot of salts deposits inside tube and outside tube --it may be coming from column overhead section ?

 

[Milutin]

Hi Petro,

As you stated earlier you find copper in your top P/A heat exchangers, if your bundle is made from carbon steel only way to have copper is deposits from column.
This deposits also can cause underdeposit corrosion on heat exchanger bundle, and in that way accelerate corrosion and deposit generation in P/A heat exchangers.

We have similar situation, from pumparaund pump naphtha P/A flow is divided in two and every flow has two heat exchangers in series. I noticed that first heat exchanger in series has much more deposits then second, and we also have tube leaks caused by corrosion only on those (first) heat exchangers. Is it similar fouling in your case?

Reasons for this in my opinion are deposits entrained from column and water. This water is not free water, naphtha from column is saturated with water and when it reaches heat exchanger cold tubes, it cools down and water becomes free water.

You mentioned process changes: installing Hi-Fi trays (high capacity) instead of ordinary valve tray, deleting down flow to tray below, installing new pumps etc… Main reason of all those changes , as I conclude, is decreasing naphtha end point, to do this you need more top P/A duty and high capacity trays (because of increased vapor – liquid traffic in that part of column). This also as process change means you decreased column top temperature. Decreased top temperature also mean higher possibility to have local free water on top tray.

Regards,

Milutin

 

[Petro0707]

The naphtha from the draw tray after pump discharge it goes to total 3 streams which are in parallel : 1st is 01 A/B exch. , 2nd is 02 A/B exch. & 3rd is hot bypass , there is no series arrangment. Over and above chloride in the deposits which we have tested in lab was less i.e. 0.09 % (may be during online water wash it got removed).Actually, our one of the purpose to kept on adding extra pump in p/a pumps was to maintain 10 C dew point margin ,yes it has leaded to increasing naphtha p/a exchanger duties too,but we kept on adding pump,eliminated external IR & kept internal IR only to achieve the dewpoint margin between dewpoint temp. of water and overhead temp. / return temperature.But hardly, we could achieve 7-8 C & not 10 C ,On and average of datasets & history , I can say that our top dew point temperature of water remains 92 C & if you see my previously posted I-MR charts of return temperature in the column,you will find that margin was at least 7-8 C min. (if you see last 8 months average) all the time.Is this figure ok ?

If you see I-MR charts Overhead temperature of the column was almost on and average 110 which is well above dew point temp of water.

I am still finding some other causes too..

 

[25362]

I may be wrong. Kindly recheck your water dew point, at a total pressure of 1.6 bara, it seems that it would be near 99-100^oC for a mol fraction of [≥]0.6.

 

[Milutin]

Hi Petro,

If your process changes were because of water dew point, does it means that you have history of such high corrosion rates on top trays? Or in other words did you experience such plugging and corrosion earlier, and what was the life of trays before process changes.

When I refer heat exchangers in series I mean situation as on the picture. Exchanger B was always more fouled then A.


Regards,

Milutin

 

[0707]

Petro0707


In my first post I have told you that Naphtha section pressure increase was an indication that maybe the 3 trays in Naphtha section were fouled with ammonia chloride deposits. NH3 as overhead neutralizer is prone to Monel because it promotes ammonium chlorides fouling and corrodes copper by forming dark blue cupric corrosion deposits.

In the past we have experienced a similar situation.


Good luck

Luis Marques

 

[25362]

Petro0707, please consider that by stopping the "external" naphtha reflux, the number of naphtha moles distilled dropped, since all this reflux would vaporize together with the normal product, and as a result the mol fraction of steam (water), as well as its partial pressure, increased.
Therefore, I still suggest to recheck the water dew point.

 

[Petro0707]

Yes, Milutin but we could find no major difference in terms of salt depositions in A or B exchanger as drawn by you , more or less salt depositions are same in both the exchangers & we calcuate dew point temperature of water by considering mole fraction of water in the overhead section ( i.e. considering partial pressure of water & partial pressure of Hydrocarb & considering total pressure )..But now we are raising doubt to ourselves whether this mehthod is correct or not because it does not take in to consideration of desublimation temperature of NH4Cl, desublimation of NH4HS etc....Can you pl. say what is the general practice...actually, we go by dew point delta temperature between water & overhead temp. or p/a return temperature and min. we keep 7-8 C delta between these temperatures.I think...should we incorporate desublimation temperature of NH4Cl , NH4HS ..? what should be the formulae of dew point temp. calculation in this case..? where can i get desublimation temperature of these components..? How can i keep soft tag in DCS of online dew point temp. monitoring as we are doing online monitoring of dew point temperature of water.?

With in almost 2 years of operation -top tray is becoming paper thin .Last time we founded the same problem in other train..how can we achieve atleast 5-6 year runlength.?

By increasing the p/a return temp. or overhead temp
of column will call for higher end point of naptha (at present 163 C & i think simulation study shows that by increasing top temp. by say from 116 to 122 C --> Naptha EP will be 170-172 C, which will create problem for platformer unit..

Should we do NH3 balance ? We are injecting aq. & gaseous NH3 in overhead line, i don't think there is a problem due to this..but yes, H2S,NH3 in crudemix...can create problem.

 

[Milutin]

Petro,

It is possible increase top P/A return temperature without increasing column top temperature. You could do this if your other (lower) pumparounds have unused capacity.

You should gradually increase this lower P/A duty on maximum capacity, top temperature controller will, suppose it is in cascade with flow controllers in front of top P/A heat exchangers, decrease flow across top P/A heat exchangers. After this you should increase top P/A return temperature set point, as result P/A return temperature go up and P/A flow go up. Final result increased top P/A return temperature.


Regards,

Milutin

 

[Petro0707]

Actually, P/A exchanger immediate o/l temp. is low 90 °C & after mixing with hot bypass final reflux temperature to column is 97-100 C which was indicated in the I-MR graph.Can you pl. suggest ionic model or NH4Cl,NH4HS desublimation temperature –how to calculate it or how to incorporate effect of these components desublimation in dew point margin..?

 

[25362]

The example by D.S.J. Jones, in his book mentioned above, gives, without external reflux, a steam mol fraction of 0.72, with (cold) external reflux it dropped to about 0.6.

 

[Petro0707]

Dear 25362,

Are you talking about the one which has shown on page no. 75 ( Section 6.9 ) in D’Jones or else? , Actually, we are using the below dew point temperature calculation formulae is it in order ?? or what will be the dew point in case of inclusion of NH4Cl,NH4HS desublimation temperature ? Can you pl. help on this (desublimation temperature graph of NH4Cl & NH4HS )

Water mol fraction :WM=(FIC23-(FIC12+FIC83+fI99))*(0.9870*1000/18)
Hydrocarbon mol fraction :HM=(FIC48*0.6470*1000/90)+(FIC36*0.6580*1000/95.4)
TM=WM+HM
MFW=WM/TM
PPW=MFW*0.968*(PIC5+1.03324)
DP=(PPW^0.25)*100

Where

FIC23 –Boot water to sour water stripper
FIC12- Fresh water overhead fin fan cooler Upstream
FIC13- Recycle water from boot water to overhead finfan cooler upstream – Generally remains closed
FIC99- Surge drum to overhead fin fan cooler upstream – Generally remains closed
FIC48- Recontact drum Hydrocarbon flow to Sat gas con
FIC36- Cold reflux – Generally remains closed
PIC5-Column top pressure
DP – Dew point temperature
PPW – Partial pressure of water

As per Crude Assay data, Eocene & Ras gharib contains more potential H2S ( as high as 4800 & 2400 ppm) whether processing these crudes can cause corrosive reactions in crude column top section ?