7 replies to this topic

[Mohsen Khodaee]

Hello everyone

Here's an attachment of the simulation of a refrigeration cycle for butane liquefaction in HYSYS. In this simulation I've used a recycle unit before each compressor. My goal is to optimize this cycle. So I've used HYSYS optimizer and my target function which I'm trying to be maximize it, is the COP of this cycle. I've defined the heater's heat flow as variables for this refrigeration cycle (3 variables) to maximize its COP. My Problem is that I don't know which method is appropriate for this goal (SQP,Mixed,FletcherRevees, QuasiNewton or Box).

Attached Files

•  Butane Liquefaction.pdf   18.2KB   58 downloads

Edited by mhsnkhd, 28 July 2016 - 06:51 AM.

[Bobby Strain]

This looks familiar.

 

Bobby

[Pilesar]

One of the nice things about simulation software is that it does not cost much to evaluate multiple options. Try a method until you find one that works. If you are habitually solving optimization problems, you will get a feel for which method works best for a particular type of setup. SQP is a generally robust method and I would start with that for most problems. If you want to learn about the mathematics behind the different optimization methods, you can find passionate champions willing to point you to their doctorate dissertations. I think such math is interesting, but most real-world employers do not pay extra for that knowledge.

[Mohsen Khodaee]

Hi dear Bobby

yes, this looks familiar because we discussed about it earlier.

Thanks dear Pilesar for your answer.

Now I would like to know in these cycle that down stream flows is mixing with upstream flows, we should use Recycle units or there is another way to do that? My problem is that when I simulate total cycle, optimization become very hard and with any change in variables the cycle could'nt converge! I've tried all of the optimization method and none of them could solve my problem! Optimization fails with this error: "Error. X value(s) out of bound."

[Bobby Strain]

And you ignored my advice to use only one recycle operation. If you set the simulation up properly you may have a chance to succeed. And I don't know what COP is.

 

Bobby

[Mohsen Khodaee]

And you ignored my advice to use only one recycle operation. If you set the simulation up properly you may have a chance to succeed. And I don't know what COP is.

 

Bobby

Thanks a lot dear Bobby.

By COP I mean Coefficient Of Performance which is the ratio of total removed heat (Q-c) over total compressors power. If I can maximize COP, then I can claim that my refrigeration cycle is working in optimized condition.

And about your advice, I tried to remove two recycles and use just one, but it's not possible. Because when down stream flows is mixing with upstream flows, I have to use Recycle unit, otherwise HYSYS simulator couldn't solve this problem and the material streams' color turn to light blue! which means it's not possible to do that. For example when the outlet of V-100 (separator unit), HP5-V Material stream is going to mix with MP2 and then enter to the 147-K-101-HP, if I don't use recycle unit, every thing goes wrong and the cycle color turns to light blue! Please let me know if I'm right or not.

Thanks in advance.

Edited by mhsnkhd, 29 July 2016 - 10:07 AM.

[Bobby Strain]

      If you give the problem some thought, you can develop a flowsheet that uses only one recycle operation, or none at all. Just consider how the system actually operates in the physical case. You should always move one step at a time. Don't try to get an answer without doing some preparation. The approach you have taken is like shooting -- "Ready, Fire, Aim". Trying to optimize the flowsheet you have will prove to be quite a task as you have discovered. When you get a proper flowsheet, you will have learned a lot about how to use a simulator. Don't just cobble together a bunch of operations that seem to work. Think it through before launch.

      In this case you have a fixed duty to condense and subcool butane. The final condition and initial condition of the butane stream is fixed. So you have only the freedom to vary the intermediate butane temperatures to minimize the compressor power. So your objective should be to minimize the compressor power requirements. A commercial operation would likely not subcool butane below its atmospheric pressure temperature. So the real operation would not use propane refrigerant, but use the butane itself and eliminate all of the heat exchange surface. But maybe your assignment is as you have proceeded.

 

Good luck.

Bobby

[Mohsen Khodaee]

Thanks for your useful advice.

 

I'll try to change my flowsheet as you said before. I would like you to know that I'm doing pinch and exergy analysis to optimize this cycle and at the same time I'm trying to optimize this cycle with HYSYS to compare my results, so in future I will ask some better questions.

Thank you for all your explanations.