22 replies to this topic

[h.eghbali]

Dear all.
HELLO.

I have an issue.

As illustrated on attached picture, liquid ethane is warmed and by thermosyphon mechanism, two phase ethane goes back to circulation drum.

Unfortunately U-type piping right after cold box is done, so it has a negative effect on thermosyphon moving and during decreasing flow of hot stream, thermosyphon flow would be stopped or oscillated.

One solution is to introduce a 1 inch pipe containing vapor ethane to reinforce thermosyphon flow during low heat transfer on cold box. What's your Idea?

Attached Files

•  20211110_092834_o5rbu7.jpg   35.21KB   0 downloads

[Bobby Strain]

Is this already installed? You offered a gas injection line size, don't tell us any more. A dimensioned sketch may help.Along with flow.

 

Looks like it was designed by a novice. There may be a lot more wrong. You may want to engage an experienced engineer to evaluate the system. And propose modifications.

 

Bobby

[Pilesar]

Theoretically, your idea should work by reducing the density of the fluid in the return line. I've used similar method in the past just long enough to establish reboiler flow in the right direction. Once the flow was established after very few minutes, the injection was closed. I just did this once using steam injection on the process side of an MEA CO2 stripper reboiler about 35 years ago and have not found opportunity to apply the technique since. Your diagram shows the injection pressure the same as the liquid ethane pressure which suggests the possibility of reverse flow in the vapor injection line. Before trying this, evaluate the changes that would take place... more heat to the system, more vapor to the drum, possibility of reverse flow, etc. Long term, you might want to evaluate the hydraulics and add additional pressure drop in the liquid ethane line -- perhaps a butterfly valve -- so that the liquid ethane will be closer to its boiling point when entering the cold box.

[breizh]

Hi,

Do you have a way to increase the height of liquid in the drum and to observe its impact on the recirculation before reducing the stream of "hot stream " in the cold box ? 

Note :I do remember having a butterfly valve at the outlet of the thermosyphon line to stabilize the loop .

 

Rq: Could it an option to install a circulation pump ? 

 

Breizh 

[h.eghbali]

Theoretically, your idea should work by reducing the density of the fluid in the return line. I've used similar method in the past just long enough to establish reboiler flow in the right direction. Once the flow was established after very few minutes, the injection was closed. I just did this once using steam injection on the process side of an MEA CO2 stripper reboiler about 35 years ago and have not found opportunity to apply the technique since. Your diagram shows the injection pressure the same as the liquid ethane pressure which suggests the possibility of reverse flow in the vapor injection line. Before trying this, evaluate the changes that would take place... more heat to the system, more vapor to the drum, possibility of reverse flow, etc. Long term, you might want to evaluate the hydraulics and add additional pressure drop in the liquid ethane line -- perhaps a butterfly valve -- so that the liquid ethane will be closer to its boiling point when entering the cold box.

 

 

Dear Pilesar.

Thanks

About reversing flow, operator should inject vapor ethane when thermosyphon moving has not been stopped yet. So injected stream could get solubilized and move with normal flow direction of two phase stream.

Adding a butterfly valve is a good idea, but we have LV which controls the liquid level in circulation drum. So I think having both butterfly valve and level control valve at the same time is meaningless.

 

https://pasteboard.co/oHwb6nFlkDTT.png

[h.eghbali]

Is this already installed? You offered a gas injection line size, don't tell us any more. A dimensioned sketch may help.Along with flow.

 

Looks like it was designed by a novice. There may be a lot more wrong. You may want to engage an experienced engineer to evaluate the system. And propose modifications.

 

Bobby

 

Dimensional sketch???

https://pasteboard.co/pQPGpdMhJDUC.png

[h.eghbali]

Hi,

Do you have a way to increase the height of liquid in the drum and to observe its impact on the recirculation before reducing the stream of "hot stream " in the cold box ? 

Note :I do remember having a butterfly valve at the outlet of the thermosyphon line to stabilize the loop .

 

Rq: Could it an option to install a circulation pump ? 

 

Breizh 

 

Dear Breizh .

Yes, we can increase liquid level in circulation drum.

About circulation pump, no we can not do it. And also this reboiler is thermosyphon and we don't want to change this mechanism.

[Bobby Strain]

What you posted is obviously not your installation!

 

Bobby

[Pilesar]

The circulation rate through the thermosiphon has nothing to do with level control of the drum! Fluid flow requires a pressure differential. In a thermosiphon exchanger, flow results when the return fluid has a low enough density relative to the cold source fluid to overcome the flow resistance caused by friction and elevation change. Assume the liquid at the pressure of the circulation drum is at its bubble point temperature.  The fluid entering the exchanger is at higher pressure than the drum because of the liquid head. Therefore the bubble point temperature of this fluid is higher because of its higher pressure.

 

Since the cold liquid entering the thermosiphon exchanger is subcooled, the reboiler must heat the subcooled liquid to the boiling point before boiling begins. Heating a subcooled liquid has a much lower heat transfer rate than boiling a liquid. When the hot stream flow is reduced too much, all these effects add up: lots of subcooled cold stream plus low heat transfer plus reduced heating duty available resulting in too little boiling. One solution: reduce the cold side flow rate so that less heat is transferred in the subcooling region and more heat is transferred in the boiling region where the transfer coefficient is better. The return fluid would have higher percentage vaporization and less density. Increasing the friction in the piping to the exchanger will reduce the pressure of the fluid entering the exchanger, bringing it closer to its bubble point temperature, while also reducing the circulating flow rate. Increasing friction by reducing the pipe size, adding an engineered orifice, or adding a valve are all legitimate. Adding a valve is nice because the final restriction is not so dependent on engineering calculations but can be determined by simple modification in the field. If a valve is in line, it would be a manual valve, set once, and never needing adjustment again. An orifice would do the same thing, but hydraulic calcs can be tricky and I prefer not to take the chance of spec'ing the wrong size. The extra pipe friction would not jump-start a stopped line, but would make it less likely that the line would stop in the first place.

[horatorres]

Increased the level in the drum to test the liquid circulation

 

 

Horacio

[breizh]

Hi ,

To support your work a few documents about the thermosyphon .

Breizh

[h.eghbali]

 

The circulation rate through the thermosiphon has nothing to do with level control of the drum! Fluid flow requires a pressure differential. In a thermosiphon exchanger, flow results when the return fluid has a low enough density relative to the cold source fluid to overcome the flow resistance caused by friction and elevation change. Assume the liquid at the pressure of the circulation drum is at its bubble point temperature.  The fluid entering the exchanger is at higher pressure than the drum because of the liquid head. Therefore the bubble point temperature of this fluid is higher because of its higher pressure.

 

Since the cold liquid entering the thermosiphon exchanger is subcooled, the reboiler must heat the subcooled liquid to the boiling point before boiling begins. Heating a subcooled liquid has a much lower heat transfer rate than boiling a liquid. When the hot stream flow is reduced too much, all these effects add up: lots of subcooled cold stream plus low heat transfer plus reduced heating duty available resulting in too little boiling. One solution: reduce the cold side flow rate so that less heat is transferred in the subcooling region and more heat is transferred in the boiling region where the transfer coefficient is better. The return fluid would have higher percentage vaporization and less density. Increasing the friction in the piping to the exchanger will reduce the pressure of the fluid entering the exchanger, bringing it closer to its bubble point temperature, while also reducing the circulating flow rate. Increasing friction by reducing the pipe size, adding an engineered orifice, or adding a valve are all legitimate. Adding a valve is nice because the final restriction is not so dependent on engineering calculations but can be determined by simple modification in the field. If a valve is in line, it would be a manual valve, set once, and never needing adjustment again. An orifice would do the same thing, but hydraulic calcs can be tricky and I prefer not to take the chance of spec'ing the wrong size. The extra pipe friction would not jump-start a stopped line, but would make it less likely that the line would stop in the first place.

 

Dear Pilesar.

Thanks for your comprehensive answer. Really impressive!

You said: "The circulation rate through the thermosiphon has nothing to do with level control of the drum!". I'm not agree with you. When circulation drum  level decrease, degree of subcooling of the fluid entering to the reboiler decrease, so heating zone decrease. So more heat transfer will occurred at boiling zone and circulation rate would increase.

All of your other statement is completely true and I'm tanksful for it.

[h.eghbali]

Hi ,

To support your work a few documents about the thermosyphon .

Breizh

 Thank you Breizh

[h.eghbali]

What you posted is obviously not your installation!

 

Bobby

Dear Bob.

You're so thoughtful.

Here is more detailed picture..

https://pasteboard.co/xQ4Foa4LYEUw.png

 

Some of the piping's are inside the coldbox! 

[Pilesar]

The detailed picture you posted contradicts your original diagram. The U-piping is designed on purpose inside the cold box, is not the problem, and should remain. Your idea of ethane vapor injection in the vapor piping is not a workable solution at all. The cold box designer should be able to explain to you how the system is supposed to operate. Is your diagram also mistaken where it shows one hot stream and one cold stream in the exchanger? You supply few details of your process, but I suspect the ethane loop is intended to serve as a refrigeration utility which only requires a correct level in the refrigerant drum. Loss of heat transfer and intermittent refrigeration flow is likely caused by temperature pinch in the exchanger as a result of operating the system outside its design parameters. If the refrigeration flow is sputtering, look to the entire system instead of the refrigeration utility. Cryogenic exchangers are tuned for specific conditions and stream duties to achieve very tight temperature approaches. You may find it useful to model this process in a steady state simulator since even a slight bow in the heat curves affects performance under these conditions.

[h.eghbali]

The detailed picture you posted contradicts your original diagram. The U-piping is designed on purpose inside the cold box, is not the problem, and should remain. Your idea of ethane vapor injection in the vapor piping is not a workable solution at all. The cold box designer should be able to explain to you how the system is supposed to operate. Is your diagram also mistaken where it shows one hot stream and one cold stream in the exchanger? You supply few details of your process, but I suspect the ethane loop is intended to serve as a refrigeration utility which only requires a correct level in the refrigerant drum. Loss of heat transfer and intermittent refrigeration flow is likely caused by temperature pinch in the exchanger as a result of operating the system outside its design parameters. If the refrigeration flow is sputtering, look to the entire system instead of the refrigeration utility. Cryogenic exchangers are tuned for specific conditions and stream duties to achieve very tight temperature approaches. You may find it useful to model this process in a steady state simulator since even a slight bow in the heat curves affects performance under these conditions.

contradicts with original one? WHY?

In both picture, reboiler outlet piping is U-Type. plz pay attention that coldbox is just an enclosed area with thermal insulation and the reboiler is inside the colbox!.

two-phase pipe out of reboiler comes down then straight and then goes up.

[Pilesar]

Your original sketch shows two streams passing through a cold box with a piping dip outside the cold box. You did not draw an exchanger inside the cold box. I don't know why the second drawing contradicts your original sketch. I assume you did not intend to mislead and you do not understand the reason the piping was designed to have a low point at the cold box interface and therefore sketched the piping inaccurately. That is why I suggested you contact the the cold box designer so they can explain how the system is supposed to function. They should have supplied an operating manual with the equipment, but I would not be surprised if the reasons for the design details were left out. Usually, the customer does not care much about the coldbox internals. The manufacturer should not mind explaining more if you want to understand the design better.

  I am not sure what more I can do to help you. I asked only one question in my previous post and you chose not to address it: 'Is your diagram also mistaken where it shows one hot stream and one cold stream in the exchanger?' The only detail you have given of your system shows one hot stream and one exchanger inside a cold box which you represent as equipment supplied for the purpose of boiling a cold ethane stream from a cold supply vessel outside the cold box. There are much easier ways of vaporizing liquid ethane if that were the true intent. I suspect you are ignoring critical system information that indicates where the true problem lies. The problem is not with the ethane system piping. You are not supplying the process information to this forum, so I suggested you simulate the process to better understand and troubleshoot it. Maybe someone else here can give you an answer you like better.

[breizh]

Hi,

Can you draw a sketch with all the piping (including inside the cold box) for us to better understand ? Accurate and complete drawings with save tons of words.

People are willing to help ,help them.

Breizh 

[h.eghbali]

Dear Pilesar and Breizh 

all data is attached. I'm sorry, but duo to proprietary regulations I couldn't upload all PFD and PID data. 

 

https://pasteboard.co/kOgNNlCQENIO.png

 

Exchanger AIM: Cooling of feed gas entering demethanizer column, using ethane refrigerant and very low temperature SGA (sales gas)

[Pilesar]

The exchanger appears to be a brazed aluminum heat exchanger with two cold streams and two hot streams. These streams do not work independently. The hot streams are coming from another multi-stream exchanger and there is at least one additional multi-stream exchanger prior to that one. The demethanizer feed system with its multiple exchangers and flash drums is very integrated and should be evaluated as a system. A steady state model of the system is difficult because of the integration. Demethanizer feed system is tough to model and will take some time. The close internal temperature approaches in the exchangers give results that are not intuited by looking at just the inlet and outlet temperatures. Building a model for design is easier than for rating an existing unit. In your case, I would first find the design basis for your plant and model that to match the design material balance. Your cracking yield will be different and the design basis may be far from how the plant currently operates, but you need to verify your model. Then you can try to change the model conditions toward your operation.

  The only problem you mentioned is the variableness in the ethane refrigeration system. What trouble does this cause? Is it only a problem seen in the refrigerant loop or do you see a problem in the feed to the demethanizer? 

  Often the hardest part of troubleshooting is determining the real problem and formulating the right questions to answer. The original post assumed there was a piping problem in the ethane system and claimed the ethane return to the drum is two phase. What evidence do you have that the return is two phases? Do you think it was designed to normally have two phases? What do you base that conclusion on? According to your recent PFD, the other cold stream leaves the exchanger at 43 C.Why wouldn't the ethane stream have that temperature also? Could ethane have two phases at that temperature and pressure? Have you noticed that both hot streams exiting this exchanger are colder than the cold ethane stream? Is the ethane stream transferring heat in just a portion of the exchanger? There is much thought put into the exchanger design and piping design inside a cold box and until you understand why the designers made the choices they did, it is presumptuous to say they were wrong.

  I have designed demethanizer feed systems. This exchanger is the trickiest one to model and design. Concentrating on the unfamiliar ethane piping is to miss the larger picture. Understand the process and figure out what the real problem is. Modeling really helps in this complicated system. There may be no one in your entire plant who really understands how this system works.

[breizh]

Hi,

Can you contact the licensor , he may provide guidance too?

Did you check the operation manual , this kind of situations are probably described under start up or shut down ?

Good luck

Breizh 

[h.eghbali]

The exchanger appears to be a brazed aluminum heat exchanger with two cold streams and two hot streams. These streams do not work independently. The hot streams are coming from another multi-stream exchanger and there is at least one additional multi-stream exchanger prior to that one. The demethanizer feed system with its multiple exchangers and flash drums is very integrated and should be evaluated as a system. A steady state model of the system is difficult because of the integration. Demethanizer feed system is tough to model and will take some time. The close internal temperature approaches in the exchangers give results that are not intuited by looking at just the inlet and outlet temperatures. Building a model for design is easier than for rating an existing unit. In your case, I would first find the design basis for your plant and model that to match the design material balance. Your cracking yield will be different and the design basis may be far from how the plant currently operates, but you need to verify your model. Then you can try to change the model conditions toward your operation.

  The only problem you mentioned is the variableness in the ethane refrigeration system. What trouble does this cause? Is it only a problem seen in the refrigerant loop or do you see a problem in the feed to the demethanizer? 

  Often the hardest part of troubleshooting is determining the real problem and formulating the right questions to answer. The original post assumed there was a piping problem in the ethane system and claimed the ethane return to the drum is two phase. What evidence do you have that the return is two phases? Do you think it was designed to normally have two phases? What do you base that conclusion on? According to your recent PFD, the other cold stream leaves the exchanger at 43 C.Why wouldn't the ethane stream have that temperature also? Could ethane have two phases at that temperature and pressure? Have you noticed that both hot streams exiting this exchanger are colder than the cold ethane stream? Is the ethane stream transferring heat in just a portion of the exchanger? There is much thought put into the exchanger design and piping design inside a cold box and until you understand why the designers made the choices they did, it is presumptuous to say they were wrong.

  I have designed demethanizer feed systems. This exchanger is the trickiest one to model and design. Concentrating on the unfamiliar ethane piping is to miss the larger picture. Understand the process and figure out what the real problem is. Modeling really helps in this complicated system. There may be no one in your entire plant who really understands how this system works.

 

 

coldbox outlet stream is two-phase. You know why? because at that Pressure and Temperature ethane exist at two phase. Also PFD data verify that. 

Real problem: By decreasing coldbox hot stream flow, two phase flow tend to oscillate duo to lowering boiling. If we decrease hot stream flow more, thermosyphon flow stops and we have to start up this exchanger, and as I've told you first, u-shaped piping would increase start up time drastically.

[Bobby Strain]

So, looks to be a designer/fabricator error. I would go to the designer/fabricator and ask that they remedy the problem. meanwhile, you can use some higher pressure ethane from the compressor as you suggested. But not the low pressure you show on your original sketch.

 

Bobby