Liquefied of propane and butane using turboexpander 1

[faisalalalwny]

i am working at C2+ fractionation plant NGL .
We fractionate the C2+ into many products including propane ,butane and c5 . Following that process, we refrigerate the C3 and C4 separately to storage temperature using close loop system include flash drum, receiver drum, compressor, and mechanical heat exchanger fin fans . Therefore,I have been doing research about turboexpander equipment and it’s advantages to replace the refrigeration close loop . your comments is highly appreciated .

 

[georgeverghese]

Figure 29-45 in Perry Chem Engg Handbook 7th edn, in the subsection on turboexpanders, has this all worked out for you.
Propane bp at 1atm = 420degR. From this graph, 2nd law thermo efficiency is about 40% at 420degR condensing temp with a typical sensible cooling internal of say 40degF for mechanical refrigeration cycle, while that for expander cycle is only about 25%. If sensible cooling interval assigned to the refrigerant is lower than 40degF, 2nd law thermo eff is even better for mechanical refrigeration.
What refrigerant do you have for propane liquifaction ?

 

[faisalalalwny]

thank you for your comments

i will one more time to explain our process in more details .

after all products are fractionated . C3 at 100 F and 260 psig stream(header) head to refrigeration Unit which consists of four main parts.

The first part called flash drum that received the C3 from C3 header to allow flashing the hot gas in order to reduce pressure and temperature . Flash drum branches the C3 into liquid and gas ,the liquid will leaves from the bottom of the drum through pump to storage tank at liquid phase -45 F and the will withdraw from the top of drum
to feed compressor to compressed the gas into hot liquid . then , the gas will discharge from the compressor to the condenser which is air finfan cooler . the condensate gas will enter the Receiver (accumulator )drum at 180 F and 260 psig. the receiver function is to maintain the drum pressure .

My research is to replace this entire system with turbo expander.

 

[georgeverghese]

Okay, so feed propane is its own open loop refrigerant. So this must be a 2 or 3 stage (?) propane compressor, with intermediate feeds from C3 flash gas drums gas as it is dropped down in pressure in stages. A subcooler downstream of the liquid C3 receiver will help to increase cycle eff.

What is the driver on this C3 compressor?

So the sensible cooling interval here is much more than 40degF. Refrigeration cycle thermo eff in this case would be much lower than 40% and will approach that of expander cycle when referring to Fig 29-45. Note that in an open loop expander cycle ( where feed C3 is the expander operating medium), isentropic eff of the expander will drop off rapidly during speed turndown - see fig 29-49 in Perry which gives some indication of this trend. So you may need to install 2 expanders in parallel if you want to reduce the effect of this drop. Such a problem does not occur with the mechanical refrigeration loop scheme you have now.

 

[faisalalalwny]

propane compressor has 6 stages attach file is similar to what we have .
electric motor with 16.7 MW drive the compressor .

what i want to know, is my idea valid or not ? if yes how is the process going to be and is there any real example world-wild ?

 

[georgeverghese]

Think you mean 6 impellers, probably in 2 stages ? Never heard of a 6stage propane compressor.

For liquifying C3, you could try a closed loop expander cycle with say 2 choices of refrigerant (a)pure ethane or (b)mixed refrigerant of propane and ethane. These are challenging research projects, since it is not just a matter of getting high thermo eff, but getting minimum life cycle cost, which involves good estimates of initial capex also. But there is no harm running some simulation cases to see what thermo eff you could get with different configurations of compressors, heat exchangers, choices of refrigerant for a closed loop expander plant.

A more realistic goal may be to see if there are ways to improve operating flexibility or increasing cycle eff, or even something simpler like troubleshooting some vexing operational issues in your existing plant. For example, can you get better recovery of C3 at your de ethaniser column by provider colder reflux at the overheads condensor, since your current condensor is operating at 100degF with cooling water only. If you have spare refrigeration capacity at the C3 compressor, you could divert this excess liquid C3 to the de C2 column oveheads condensor. This method is used in many plants.

 

[georgeverghese]

Error in my last post:
"since your current condensor is operating at 100degF with cooling water only."
This couldnt be; this deC2 column ovhd condensor must be running on propane cooling, so see if more propane is available for lower reflux temp, assuming there is some way of decreasing ethane slip into deC2 column bottoms product leading to deC3 column.

 

[faisalalalwny]

To clear things up , I think we are not in the same direction . This is a simple draw of our process


To clear things for me again I will ask questions if turboexpander could replace that system or not .
Thank for being patient.

 

[georgeverghese]

So you have a 2 stage C3 compressor here. No subcooler shown. There may be more opportunities for better eff during turndown operations. With flash drum operating at 15psig, there must be plenty of boiloff gas being generated at 1atm for chilled propane at the 1atm flash drum. Recompression power demand of recovery of this 1atm flash gas back to interstage drum pressure will lower thermo eff further.

This scheme does not show a side bleed to the deC2 column for cooling at the ovhd condensor, so what is the coolant at that condensor ? Propane recovery ratio at that column would be quite low if you run that condensor on cooling water.

Of course you can replace this unit with a closed loop turbo expander cycle, but what thermal eff can you get ? Key factors to take into account:
a)Choice of refrigerant
b)Cold end temp approach for chiller heat exchangers - for compact plate full countercurrent flow HX, you can get down to say 2degC temp approach (1degC temp approach may not be practical with non licensed design heat exchangers). Best choice in the public domain appears to be PCHE - printed circuit heat exchangers - one vendor is Heatric, USA. The closer the temp approach at the cold end, the bigger the HX gets.

When compared to the current scheme with 2 stage C3 compressor, you can probably get better thermo eff with a closed loop expander cycle easily. But with a 3stage C3 compressor, you can further chill down the product propane to say -30degF at an LLP flash drum (operating at 3-5psig say) and increase overall thermo eff. Getting a better eff than this 3stage C3 chilling option with closed loop expander is probably worth studying.

 

[faisalalalwny]

If possible to show the idea in drawing

 

[georgeverghese]

For improvised 3 stage mechanical refrigeration, see GPSA in the chapter on refrigeration. So in your case, would suggest blending the feed C3 from deC3 column overheads into compressor 3rd stage upstream of discharge cooler / condensor. And withdraw chilled propane liquid at 1st stage flash drum exit. No need for the kettle type HX implied in the GPSA diagram.

Where gas feed pressure is low (as in this case), it doesnt make much sense to run an expander cycle for liquifaction of C3 as mentioned earlier. This is because ethane has to be recompressed back up to HP conditions in the closed loop after the expander, and this compression will take up more power than compression of propane, since ethane is lighter than propane. The author of this section in Perry further says that expander cycles are better at refrigeration than mechanical refrigeration cycles at final condensing temp of < -100degF or -73degC.

Would suggest redirecting your research into improvising the existing 2 stage C3 mechanical refrigeration cycle. You still havent said what is used as coolant at the existing deC2 column ovhd.

 

[faisalalalwny]

what is used as coolant at the existing deC2 column ovhd.?Acually we dont have in our unit dec2 column but we are only have C3 liquefaction (change the phase of C3 from hot gas to liquid storage tempreture at -45 F) and C4 liquefaction (change the phase of C4 from hot gas to liquid storage tempreture at 25 F) in order to ship it with large quantities of barrels at 300000 barrels for each ship are and that manly our unit function .

deC2 column ovhd is exist in the fractionation unit not our unit and they use propane from(reflux drum )
ovhd.

 

[georgeverghese]

Okay, good of you to provide that info.

I can see one easy way to improve operations of your C3 refrigeration unit with the use of turbo expander :-

Feed hot propane from frac unit is at 260psig, while interstage flash drum is operating at around 80psig(I suspect - pls confirm). So there is a large 160psi pressure drop which is wasted energy here. You can install an expander to perform this pressure drop to replace the existing pressure letdown valve. This will cool the feed gas more and reduce the intermediate flash drum vapor loading on the 2nd stage of the C3 compressor. To maximise on this cooling, it would be preferable to precool this gas before the expander - options for precooling are, from what I can see:
Option 1: To precool this gas against intermediate flash drum off gas going to compressor at a HX. And load the expander on the brake side by compressing flash gas downstream of the precooler HX.
Option 2: This option is a bit more difficult and simulations are required to check viability : First, flash the 15psig propane product at a new LLP flash drum down to say 5psig. Use this LLP drum flash gas to precool the 260psig / 100degF feed C3. Install a pump to evacuate the LLP propane liquid to the offsite C3 tanks. Load the expander on the brake side by compressing this 5psig LLP flash gas up to 15psig to feed the existing 15psig flash drum. This option has the added benefit of reducing boil off gas that will need to be recovered / recycled at the C3 tank also.
In both options, install a V/L knockout drum at expander HP side to remove any liquids. Dump any liquids collected at this KOD into intermediate flash drum on level control.

It is likely that you may get 2 phase vapor / liquid at the expander exit. Perry Chem Engg Handbook says, in the section on expanders :"The permissible liquid condensation in the expanding stream varies with discharge pressure; it may be 50 weight percent or higher in the discharge, provided the turboexpander has been specially designed to handle condensation." Check with expander vendor:- the more liquid it can permit at expander exit, the better.

Both these options have the effect of reducing compression power demand at the C3 compressor, essentially by reducing the flash gas generated at the intermediate 80psig flash drum. This will then allow more feed propane from frac unit to be handled at this liquifaction unit at no extra power demand.

 

[georgeverghese]

Error at option 1 precooling scheme
"And load the expander on the brake side by compressing flash gas downstream of the precooler HX."
This is not a good idea - compressing this flash gas at expander brakeside recompressor will add superheat to this gas. Unless you know of some place where this energy stream may be put to good use, I would suggest using an electrical brake generator; float it on the low voltage (460V?) bus.

 

[georgeverghese]

There is another way to configure option 1 which will make use of expander brake side at this unit :
Operate a subcooler kettle HX for liquid propane exiting the condensor-receiver vessel. Refrigerant for this subcooler will be C3 propane liquid taken from downstream of subcooler. Flash gas from this kettle can be used to precool 260psig/100degF feed propane at precooler. Then recompress this 80psig flash gas through expander brake side recompressor directly into main C3 compressor final discharge.
Run these options on a simulator to see which one is feasible or better.

 

[faisalalalwny]

sorry to stop replying to your comments .. so as far as I realize from your examples and suggested solutions that we can't use turboexpander equipment alone to replace liquified propane using the mechanical refrigeration.

 

[georgeverghese]

Turbo expander slone wont be enough, you need some KOD and heat exchangers. But there is ample opportunity for power savings here to justify the additional equipment.
At the least, the subcooler alone will save you a few percent in compression power.

 

[MJCronin]

Only 379 more questions before we get the whole picture ... Press on fellow Pilgrims !!

Oh, wait ...20 year old Fiasllalamala will have to take another THREE MONTH break before he responds

(family issues, ... of course)

Stay tuned, my friends ... stay tuned !!!

Be ready to answer ALL QUESTIONS within 24 hours .... Cheers !!!

MJCronin
Sr. Process Engineer

 

[faisalalalwny]

Thank you, much appreciated your support and i will keep digging into the subject until will come later with simulation for the solutions