Gamma Scanning 6

[AndreChE]

Dear all

Our C2 and C3 Splitter are with very high DP.

For the C2 Splitter, we have checked all the pressure profile along the column and we know in which section the flooding is located. We were suspecting ice/hydrates but we collected greenoil in the bottom level indicator nozzle.
The column was revamped with new trays, more capacity. At the moment it is distilling onspec bottom and top products just with a very high DP than expected.

For the two C3 Splitters, the situation is diferent. THe columns were not changed. They were hydrotested and closed.
When we started up the plant we experienced very high DP along both columns, with values in which was not possible to distill in the past. However, the situation still remains and both top and bottoms are onspec. Our maximum DP was 700/800 mbar and now we have 1200/1300mbar for 178 trays in each column.

We will perform gamma scanning for these 3 columns. My question are:

- Which conditions should we put the columns for the scan? Max possible load? We already have very high dp but they are distilling. High/low reflux? High/low reboil?

- Were can I learn more deeply how to analyse gamma scan plots? I have read everything in Kister's books and some small papers but I feel I need to study deeper in order to fully understand quickly those plots? Any papers or articles on the net or journals? Any specific book or documentation? The tests will be performed next week.

Thank you all
AndreChE

 

[jamesbanda]

This is my advice.

I assume you are using a scanning service provider? they should have knowledge on interpreting the plots.

I have used tru tech and traceco UK/US. Both have well trained staff at understanding the scans - they are your experts.

i dont know what you mean by green oil (i work in the chemical business) i assume out of spec product?

Here is some advice I have for you.

1. Get detailed drawings available (to scale). The scanning companys should be able to "superimpose" the GA diagrams next to the scans.

2. Contact the tray manufacturer and discuss the symptoms - i've found them generally very helpful.

3. Tru tech have artiles on their web site.

4. Kisters troubleshooting book maybe a good reference for you. I've read many chapters lots of learning.

5. I assume (you probably) have checked the DP's for fouling / freezing and checked you have not changed operating pressure ie the DP is right just lower operating pressure.

6. I would discuss with tru tech/ traceco/ other but my thoughts would be.
> Scan with the high DP. you dont know if it is one tray or 10 with high DP as it is measured over many trays.
> Scan with lower DP, scan at acceptable DP.
> Get a simulation model of the column / and DP calcs for the trays and determine what you expect the trays to do. how far is the DP out 5, 10, 50 100%. What V/L is the flood point.

> Do you have a dry scan of the columns ? (ie before startup). It is worth every penny for a restart project.

Hopefully i have been of help

Jim.

 

[pmover]

i'll second jamesbanda recommendations as i have witnessed first hand the skills, knowledge, and capabilities of tru tech. also, as stated, involving the tray mfg is benefical as well.
the results of the scan are detailed and useful in determining where problems exist within the tower.
have tower design data and process conditions available for the techs as well.
good luck!
-pmover

 

[jamiegla1]

I used to work at TruTec (which is now Tracerco) for several years. I would agree with both above...scan at your problem conditions. You dont know which tray the flooding originates, if it is total liquid flood, or jet flooding. The scan will be able to differentiate between jet and total flood. If you suspect that a large section of the tower is flooded, you may want to lower rates a bit so that the flood doesnt "mask" the problem tray(s). If it's within your budget, scan at normal conditions and again during an intentional upset.

Also get a baseline scan the next time you go into it and clean or change the trays. It wont be a "dry" scan (you dont actually see the trays on a scan, only the liquid level). This will be invaluable for troubleshooting in the future.

What type of trays did you put in it? How many passes?

 

[AndreChE]

At the moment we don't have "normal" conditions.
We just have very high DP and products onspec!! Distillation is running above expected pressure drop.

In the past, for the C3 Splitters, with 800mbar we were not able to distill. Now we have 1200 mbar and the measurement is correct.

For the C2 we have localized flooding, we just don't know why (ice or fouling), but is is mainly in the feed section. Other two sections are flooded to and we want to check why.

For the C3 Splitter we have 4 pass trays (unmodified) and 2 pass trays in the C2 Splitter (modified). We had two passes before too, just more capacity trays.

 

[jamiegla1]

In your case dont think id upset the C3 splitter. Just scan at normal conditions and make a judgement call after you see the first scan. You might consider scanning more than one pass in this tower. Just a thought but are you sure the instrumentation is correct?

For the C2 splitter, is it possible that you are feeding a "subcooled" liquid (not truly subcooled but too cold for enough heat xfer on the trays). this will show up on a scan and the engineer or tech will point it out. also, maybe you carried some debris in with the feed and it settled in a downcomer somewhere. they can also scan downcomers if you need them to.

Which state are you in?

 

[jamesbanda]

Andrew,

Did you buy the new trays from Koch or sulzer or other. My experience of Koch/Suzler is that their technical engineerings are top draw. I am sure other manufacturers of trays are similarly good. I would ask if they can help. They should be able to give you guidelines for normal DP's, flood etc...the other thing is that they have problably - maybe seen this kind of thing in the past. even if they require a payment it is probably <1% of your lost profit per day..

Anyother factor..I will be totally honisty. I struggle to understand complex column problems unless i have a GA diagram to scale and write down the symptions on where we think they are..

Some more thoughts.

I note you are in spec.
Have you validated your feed composition temps/pressure, species. I am not a refinary bod so is there a possibility of a trapped component causing DP build ?
Getting a true mass balance is always stated as ease..it can be hard work to make sure it is reliabible but you have probably done this already..

 

[0707]

I suppose for the gamma scanning you should have the columns at the actual conditions.

We have already performed gamma scanning and also TruTec. For me TruTec is a more accurate technique.

In the C2 Splitter if you had perforated trays and you change to high capacity valve trays it is normal that you have now a greater DP.

For the two C3 Splitters maybe you have the trays fouled with polymers and the ascendant light vapours have no force to unplug the tray holes or push the valves of the trays.


Regards

luis

 

[AndreChE]

We have performed gamma scans for C2 and C3 Splitters.

C2 splitter was revamped from sieve to valve trays, however expected DP is much lower than actual.

C3 splitter was not retrayed.

First preliminary results:

C2 Splitter: foam from feed tray to the top (75 of 100 trays). Bottom sections is clear and operating well

C3 Splitters: foam and liquid flooding in downcomers. Probably fouling in the sieves giving high resistance to vapour flow. We have found a particular strange thing in the first splitter: 300mm of liquid between the distributor and the top tray --> severe fouling?

What could cause this foam in the C2 Splitter? Does greenoil induce foam?

For the C3 splitters, corrosion and iron oxide mud (or other) can cause this since the column was emptied after hydrotest very quickly and with the top manhole open and then closed until plant startup.

Anti-foam for c2 splitter? Does anyone know a good one? Efective experience?

Regards,
AndreChE

 

[0707]

Foam is not unusual in C2 splitters during start up sets, until unit stabilization.
Same times it is helpful to use antifoams.

For the C3 splitters we have already experience valve tray plugging after a turnaround.

If this kind of vessels stay open for long periods of days until start up the iron oxide scale stuck around valve trays creating hard oxide layers which will impeach the valve tilt on tray holes and as result trays got dry because ascendant light vapours have no enough force to brake iron oxide scale around tray holes for that reason the column rectification is minor and all the liquid will stay on the bottom of the columns.

Good luck

Luis

 

[AndreChE]

Yes, I believe on that too. After column opening, were you able to see that? valves or sieves blocked by hard iron oxide?

For the revamped column I have Sulzer discussing the subject and taking a look for the gamma scans.

Since the active area is higher than with the sieve trays could be possible to have much more aerated liquid in the tray. This only happens above feed, which is most of the time, 100% gas.

However, Sulzer design must have checked this issue in the design phase. Even in my simulations in Aspen Hysys, this several behaviour is not pointed.
All other revamped columns are working fine.
We have seen greenoil in the tower bottom level indicator nozzle. Probably during startup it has passed through the cold section. The upstream tower is ok so...it's strange to believe on that.

 

[0707]

“Yes, I believe on that too. After column opening, were you able to see that? Valves or sieves blocked by hard iron oxide?”


Yes you have to open all vessel visit doors to see where the valve tray sections are more plugged, and then you have to hammer all the valve trays from the bottom of the trays until you will be confident that all of them tilt freely in tray holes. It is a time consuming job but I don’t see another way to do it.

If by sampling you would suspect of polymer gums, in complement of my reply if you have increase the production of reactive diolefine as they have weak vaporization of the feed and you had to increase temperature, the alteration in the concentration of butadiene leads to its polymerisation. Iron depositions promote polymerisation.

1º - Bad separation of the water leads to the increase of the corrosion in stripper and in the debutaniser. The corrosion products (iron sulphides) are accumulated in the low points of reboilers.

2º - the increase of the temperatures in reboiler above of what is usual, leads to the formation of gums. These gums supply the "glue" to fix the corrosion products. To delay the formation of "fouling", others have used with some success, on line dispersants and inhibitors of gums formation.

If this is your case and you want experiment to avoid opening the columns, you can try to inject on-line dispersants and see if the things got better, if it doesn’t work you have no other way but to open the columns and unplug the vessel trays manually.

Good Luck

luis

 

[jamiegla1]

do you have any amines in the C3 splitter? maybe injected upstream to control corrosion? i believe they are notorious for causing foam. the foam can also partially block your downcomers, causing the trays to flood.

 

[STI666]

Scanning Technologies provides Gamma Scanning services as well as Tracerco/TruTec. I would suggest Gamma Scans be performed across all of the active areas of the trays as one section may be giving you problems and the other side not.

To evaluate antifoam use, Gamma Scanning can be performed in real time. Start the Gamma Scan and find the level of the foam. Then start adding antifoam and scan again to see the new foam level. This is a useful technique for determining how much chemical is needed. Depending on where you are located a large cost of the Gamma Scanning may be due to mobilization. Once the crews are on location additional scans are relatively inexpensive in our experience with Scanning Technologies.