[evan]

It is hard to say where the COS is coming from without seeing the lab analysis of the various streams and know exactly where, when, and it what rates is is injected……have you tried sampling the LPG at various points to see where the COS is coming from or where it maybe removed to gauge the effectiveness of the LPG treating?

[evan]

Additional comments from Chris Dean…..

Thanks Hedewandro for a very good question,

As previously mention the amount of COS produced is based on the feedstock. With only amine treatment for the LPG stream and/or PP Stream, it is difficult to reduce COS levels significantly or even at all. This could be confirmed by a SRU SME or from an amine supplier or treating licensor. To reduce COS significantly, catalytic treatment beds are used to produce polymer grade propylene as feedstock to polypropylene units.

You mention in the question that Coker LPG is sent to a propylene unit. Is this unit a PP splitter or does it have other treating facilities to produce polymer grade propylene? Most refineries will produce refinery grade propylene and some depending on location will produce chemical grade.

Just for information the typical COS spec is 0.5-1ppm for polymer grade propylene. Chemical and Refinery grades are negotiable due to the customer but COS does affect total sulfur.
Hope this helps

[evan]

Determining oxygen in resid is tricky. I’m not aware of any simple tests but FTIR and neutron activation analysis have been used in labs in research papers that I have read about ashphaltene molecules.

Regarding AF, 100k cSt is middle of the road in terms of MW. You might consider 300k or 600k cSt.

Purge, 6 volumes may help but I would sample to check and determine if the direction change is effective. Come to think of it, I would sample before and after any change to measure its effectiveness.

If you make a change, please let us know on the thread if it works!

[evan]

Hello

I believe the source of most of the COS is derived from the oxygen and sulfur species in the feed. Has something changed in the feed?

Other sources may include
+ O2 from the open drum. This can be reduced by increasing your purge duration, minimum 6 drum volumes.
+ PDMS from antifoam has oxygen species that can be converted into COS. Antifoam injection practices can be optimized to use less and higher MW antifoam can be used to use less as well.

[evan]

I do not know of a test method but I will share this thread with some friends at the chemical vendor companies in the hopes that they can help you.

[evan]

Do you have orifice or wedge? I’m confused by your last post.

If you have orifice plates,
+ continuous purging the taps
+ orienting the taps at 2 and 10 o’clock (free draining)
+ eccentric orifice plates
could help

Assymetry, seems possible. Coke fines could preferentially build up in 1 leg. Can’t say that I have seen this before but I think almost all the heaters I have experience with have symmetric feed pass piping.

[evan]

Hello

I’m not sure I understand your question. Are you looking for an estimate of how much APS to inject? Or something to do with the chemical composition itself?

[evan]

Hello

Pass flow has seen many evolutions over the years. But recently, we have been standardizing around
1. Wedge meters, large taps (2″ or more) close coupled, with very, very good insulation
2. Ultrasonic, inline. live vibration needs to be considered

Very, very rarely, we see coriolis in this service. Its the best but very expensive.

[evan]

Fransisco, thanks for adding to the conversation.

I wanted to add that some refiners have been seeing increases of 3-5degC on drum inlet temperatures by going to a high tech insulation from Aerogel. Yes, most of the dT is from endothermic reactions but insulation matter too.

[evan]

Seems a bit high dT but not completely out of line.

What is the velocity in the line?
Have you seen any increase in pressure drop in the transfer line for coking?
When was the last time you cleaned the transfer line? Last TAR was when?
Is the insulation in good condition?

[evan]

Bsudhakar – are you still operating a fuel oil fired coker heater? There are not many that I have encountered these days?

High heat values and burner/draft issues make hot spots more likely with these types of heaters. All of the causes you mention can lead to hot spots and accelerated creep rates. If you don’t have a TMT there or have good visibility with IR camera, you might never know there was hot spot there. We always advocated for more TMT’s and better peep doors for just this reason.

[evan]

Bsudhakar, I could not agree more.

+ Down gauging the AF pump is important to really know your dosage.
+ Measuring Si in Naphtha and LCGO is important. >20 ppm could cause downstream catalyst issues.
+ Consider injecting only carrier early in the foam observation, then switch to carrier+AF later in the cycle
+ Use continuous level (if you have them) to optimize dosage timing.

[evan]

Temperature drop is a function of…
+ velocity, distance, diameter, etc
+ reactivity (parafinic vs naphthenic vs asphaltenes/ccr etc)
+ Coil outlet temp, COT
+ Measurement location (not all COT and drum inlet Ti’s are in the same place)

That being said…. typically we expect 15-25degF drop in the transfer line.

What are you seeing?

[evan]

Hello

dT from the drum vapor quenching is controlled a number of ways but is typically 20-50F. The dT after quenching to the fractionator inlet is typically quite small. The flow is 5-10% of the drum vapor to achieve that dT. You would need to back calculate that on a unit feed basis.

Some of the 5-10% quench oil is recovered as product and does not end up a recycle but some does. The exact amounts vary but 3-5% TPR can still be achieved with the quench values.

Any chance I can meet the coking committee? The coker road show is in Louisiana right now….http://coking.local/roadshow/

[evan]

Hello –

I have participated in some vibration harmonic analysis of some units with severe vibration issues. One example had 1 hz period with a magnitude of deflection that exceeded 20 cm in some piping. The issues come and go but seem to be related to a number of issues…..

+ feed composition and vaporization rate in the transfer line
+ heater COT
+ Angle of inlet nozzle
+ Random channeling in the coke bed

Are you able to reproduce the problem with every drum or does it only occur on some drums?
Do you have skin TI around the coke drum to monitor for differential heating?

I’m not sure I understand what you mean about feed rate effects on 1 drum during decoking. I believe you only have a 2 drum unit at HDO so this should always be the case.

Have you considered the Deltavalve center feed device. If the problems are severe and consistent, this would be a good solution for your plant.

[Author]

Posts