6 replies to this topic

[mje03]

Greetings, 

I'm currently doing an internship at a refinery. One of my projects is to evaluate the cost/benefits of bringing back the sponge absorber within the rheniformer/platformer back to service. The sponge absorber's purpose is to absorb some more propane, butanes and some C5+s from the recycle hydrogen stream produced by catalytic reforming (entering near bottom of the column), using lean oil entering the top of the column. Around 10 years ago, they decided to shut down the sponge absorber because they were losing the lighter ends composition in the lean oil to the overhead of the column. 

I have looked at the historical data (L,V, and hydrocarbon compositions coming in and out of the column) and confirmed that the problem existed. However, I'm stuck on not knowing what the next steps should be like.

Any guidance towards solving the problem is much appreciated.

 

[Pilesar]

I see two possible paths. The first is to use the historical data to define the unit performance. You would first assume the performance today would be identical to history. You might build a simulation model to help determine energy usage. Then perform an economic analysis of the process using current pricing. This would be the easier path. 

  The second path would be to model the process with computer simulation to match the historical data. Then adjust the model parameters (sponge oil composition, temperatures, pressures, etc) to theoretically optimize the process. Then perform economic analysis on the optimized results.

  Both paths might be taken in series if there is time. For an internship, the second path may be beyond a reasonable expectation since your time and experience are limited. I have performed software models of similar sponge absorbers. Slight composition changes can have a great effect. Historical data is often incomplete for some critical piece of information and in this instance could not be measured directly since the unit is out of service. I would not give the modelling optimizing task to an intern because of the risk the project would not reach a satisfactory conclusion.

[mje03]

Hi there, 

Thanks for the suggestions, that was really helpful. My initial thoughts were trying to figure out if the temperature of the recycle H2 gas entering the sponge absorber was too hot or was it the lean oil feed that was too hot. Are there any ways you can think of to determine the root cause analysis? 

As for the two suggestions, for the first one, which process simulation soft ware do you think I should use based on your experience?

For the second suggestion, it is true that there are incomplete pieces of historical data. It seems like there are no composition analysis of the lean oil coming in and the product oil coming out at the bottoms. If I manage to find the historical data of the composition of these two streams mentioned, what steps do you recommend for further progress? Do I build a model similar to the first suggestion but adjust the parameters to optimize the process?

For context, this is my first internship and I am in my junior year of college. The project not reaching a satisfactory conclusion would not be a problem as long as I can document the progress and hence it can be transferred to the next intern. 

Much appreaciated.

Edited by mje03, 30 October 2023 - 09:01 PM.

[breizh]

Hi,

You may want to consult this website:

http://www.eptq.com/...af-4fc137520abf

 

Breizh 

[Bobby Strain]

Use the simulation software provided by your employer.

 

Bobby

[Pilesar]

It seems like there are no composition analysis of the lean oil coming in and the product oil coming out at the bottoms.

I am not surprised the composition information is missing. I understand your desire to forge ahead and get an answer. The more data you make up, the more imaginary your conclusions. Before optimizing a model, it should first be tuned to replicate known results if possible. You will need to find out what variables can be changed in the field and then test their effect on the calculations. Review the P&ID and the DCS screen for the unit to understand what variables can be controlled. Find an operator from when the unit was running and learn what the problems were and what was tried to fix them. Your supervisor should be able to supply guidance. Understand before you start what engineering deliverable would be acceptable. In my opinion, it is better to complete a smaller scope than leave an incomplete mess for another intern.

[shvet1]

 I'm stuck on not knowing what the next steps should be like.

 

modelling => report => refurbishing => test run => report => remodelling => report => minor changes of process => test run => report => operation

 

Start with initial design of this part of process - how it was intended, how designed, how operated, what changes made by start-up team and what by operation team, how controlled and how reported. Then an action plan / road map for your boss with a schedule and resources / persons involved / required and risks and benefits at each stage.

Edited by shvet1, 30 October 2023 - 11:46 PM.