9 replies to this topic

[gham]

The refrigeration system in our plant is a closed-loop system where it uses compressed C3 to partially condense C2 overheads for the De-C2 reflux.

Receiver drum stores the refrigerant drum at 250-300 psig and 120-140 F. C3 refrigerant flows from the receiver drum to refrigernat economizer (shell and tube heat exchanger) where it exchanges heat with C2 product. C2 gets heated from 42 F to 120 F whereas C3 refrigerant gets cooled to 112 F.

C3 then flows through Level controller valve (LCV) to refrigerant flash drum which is operating at 75 psig. There is a big pressure drop a cross the LCV which causes the hot gasses to flash inside the flash drum and cools it about 42 F. The gas leaves the flash drum through Pressure control valve which maintains the flash drum at 75 psig. The gas then goes to knockout drum before the compressor. The liquid C3 refrigerant, on the other hand, flows from the bottom of the flash drum at 42 F through LCVs to De-C2 overhead condensers (shell & tube heat exchangers)

As C3 refrigerant partially condenses c2 overheads, the temperature exchange vaporizes the refrigernat C3. C3 vapors flows from the overhead condensers, combines with the flashed refrigerant vapor from the flash drum, and goes to the knockout drum and then to the compressor.

The compressor takes suction from the knockout drum and compresses the C3 refrigerant vapor to 240-300 psig and 160-185F. The compressor is a four-stage centrifugal compressor driven by 14,012 HP Deleval Stork letdown (superheated ) 600 steam turbine.

Compressed C3 vapor flows from the compressor to the refrigerant condenser which cools the vapor to about 120 F. Now liquid C3 refrigerant flows to the receiver drum and the process begins again..

My questions are the following:
1- what is the optimum pressure of the flash drum?or how to calculate it.
2- what are the feasible options that I have in order to reduce the steam consumption? or how can I optimize this refrigeration process or how can I reduce the steam consumption or increase the process performance.

I appreciate in advance your help and your interest in answering my question.

[Art Montemayor]

gham:

This is a very good, written process description of a single-stage Propane refrigeration cycle. I have sketched the cycle, as described, on an Excel Spread sheet and I have the following comments:

1) The "optimum" pressure you require in the Flash Drum is not really an optimum figure as I understand your cycle. Rather, the pressure you maintain in the Flash Drum is what is required by -
a) the need to maintain a positive, gauge pressure in the compressor's suction in order to ensure that no partial vacuum is created - even momentarily. This is to ensure that you do not suck in atmospheric air (& it's inherent Oxygen content) into the gaseous Propane stream. The result would be a potential disasterous explosion. Normally, approximately a minimum of 10-15 psig is kept at the compressor suction to ensure this not happening.
B) the pressure in the Flash Drum is also fixed by how low a temperature you need to have the resultant flash liquid at. In other words, the lower the Flash Drum pressure, the lower the temperature the contained saturated Propane liquid will be at and the Lower the resultant temperature at the downstream evaporator(s). If you need this lower temperature, then go for it. However, you will require more horsepower because of the higher delta pressure range you will operate in.

c) the lower the pressure in the Flash Drum, the higher the specific vapor volume will be and the more capacity required from your compressor for the load. You haven't said if you are going to maintain the same evaporator load(s) or not. Pehaps you are seeking more capacity?

2) You have erroneously stated: "C3 then flows through Level controller valve (LCV) to refrigerant flash drum which is operating at 75 psig. There is a big pressure drop a cross the LCV which causes the hot gasses to flash inside the flash drum and cools it about 42 F." This can't be true. The Propane coming from the Propane Receiver must be in the liquid state and is further subcooled by the ethane heat exchanger between the Receiver and the Flash Drum. The reason for the ethane subcooler is to ensure that the Propane expanded into the Flash Drum is liquid and subcooled in order to minimize the flash vapor produced - giving the cycle a higher efficiency at the expense of heating up the ethane.

3) You also fail to state where the vaporized Propane vapors are routed to specifically. I believe that the Propane vapors from your evaporator(s) is going directly to the vapor space in the Flash Drum and joing the flash vapors there to be controlled by the back pressure controller on the Flash Drum prior to entering the compressor suction knock out drum.

4) There is an obvious thermodynamic error in stating that the Flash Drum is at 75 psig and the 112 oF liquid from the economizer is cooled down by 42 oF (down to 70 oF) by the expansion into the Flash Drum. This cannot be. If the Flash drum is maintained at 75 psig and the contents is saturated Propane (which it must be, in order to maintain the cycle) then the temperature inside the Flash Drum must be 115 oF. In order to have 42 oF saturated liquid Propane inside the Flash Drum, you must maintain it at a pressure of 27 psia (or approximately 12 psig). This would make economical and practical sense. Can you please clarify this important point in your basic data?

If you can handle it, you could increase the refrigeration effect of the cycle and also reduce the compressor horsepower by subcooling the liquid Propane into the Flash Drum expansion valve even lower -- perhaps as low as 80 oF if you had a water evaporation cooler. I suspect you are in an arid location and water cooling is dear or expensive. Therefore, you may have other options in your process, such as waste, low pressure steam with which you could produce really cold water (50 oF) for cooling by simply using a jet ejector for cooling.

Without more basic data or a scope of work, I can't produce more specific recommendations or comments, so I'll close here. Basically, you are not wasting energy when you haven't the natural means to operate at a colder level with respect to your Propane Receiver. If you are in a hotter environment, it will cost more horsepower to produce the same refrigeration - there is nothing you can do about that. However, you can make judicious, smart engineering use of normally waste process streams to reduce your load horsepower demand.

I hope the above helps out in your analysis.

Art Montemayor
Spring, TX

[gham]

Art Montemayor:

Thank you for your comments and help. I just would like to answer some of your comments that you raised up through your explanation:

1- In addition to what you mentioned about the pressure of the flash drum, it was actually set in order to reduce the size of the vaporizers by reducing the refrigerant vapor traffic and making them more thermally efficient. With the design pressure of 68.2 psig (sorry not 75 psig), 33% of the flow going into the flash drum will be vaporized.

2. As you stated, the propane coming from the propane receiver must be in the liquid phase which it is. (I miss typed the sentence).

3. The vaporized propane is combined with the propane vapor coming from the flash drum and the combined stream goes to the knock-out drum (I have attached the process flow diagram that I made in Hysis in order to optimize this process).

4. There is no any thrmodynamic error in what I mentioned about the flash drum and its operating conditions. The liquid propance (at 290 psig, 112 F) goes through the level control valve where a pressure drop occurs and liquid propane flashes into the flash drum where the pressure is set at 68 psig and cooled to 43 F (not 70 F). At 68 psig and 43 F, you should have saturated propane.


What I am really looking for is reducing the compressor horsepower. As far as utilizing waste water or other streams to subcool the liquid propane is just not feasible in our environment.

Is it possible to increase the pressure of the flash drum to a point where you can still achieve your refrigeration requirments.

Thanks again for your valuable comments.

Attached Files

•  Refrigeration_process_flow.doc   48.5KB   615 downloads

[Art Montemayor]

Gham:

Thank you for correcting my Thermo data error! You are absolutely correct about the saturated thermo values for Propane. Before I depart for our family's 4th of July celebration, I want to make sure you receive this. I always use the handy NIST thermodynamic data base available at:

http://webbook.nist....hemistry/fluid/

but this time, in my haste I selected Propane as the fluid and didn't click off on the selection. This is a programming bug on this database and I forgot about it and proceeded to scroll down with my right mouse button to start the program. Another bug this program has is that the tabulated results you print out do not state nor title the name of the fluid in question. As a result, I erroneously got the results for Isobutane, thinking I was downloading Propane values into my spreadsheet. I apologize for my hasty error.

I have the data in an Excel workbook, complete with flow diagram and calculations and would like to share it with you on how I perceive the process could be improved, but I'm trying to find out how you can submit attachments on this forum and I'm still learning. If you contact me at my email, I can send you the workbook directly to you. If I knew your specific evaporator refrigeration load, I could generate specific numbers for your case. That way, you could evaluate various methods of improvements to the process and identify their absolute economic worth for your case.

Art Montemayor

[gham]

Art Montemayor,

Thank you for your interest and help in answering my question.

The evaporator refrigerant load is 528,800 Ib/hr (about 20,686,000 Btu/hr) to condense about 1,063,000 of C2 (from 46.8F to 41.2F).

You can send your response to the following email alghamdi78@hotmail.com

Thanks once again for your help and Happy 4th of July.

[Art Montemayor]

gham:

I have found I cannot attach an Excel workbook to this posting. So I am sending you my calculations in a separate email.

You will note the following:

1) Your figure of 528,800 lb/hr of Propane evaporator feed doesn't seem to fall into the 20, 686, 000 Btu/hr refrigeration load you cite. For example, the latent heat of evaporation for Propane at 68 psig & 43 oF is 157 btu/lb. Therefore, the refrigeration duty on the evaporator is:

Q = (157) (528,800) = 83,021,600 Btu/hr

2) I don't believe you want to raise the Flash Drum pressure; this will lower the driving force on the evaporator since the Flash liquid temperature will go up correspondingly with the pressure. The end result would be that you get less refrigeration effect. Your evaporator may not be able to meet the duty.

3) Subcooling the liquid Propane feed to the Flash drum yields a dividend - as is often the case in almost all refrigeration cycles. What is being done here is that you are trying to make the most of low-cost cooling at the higer levels prior to getting into the more costly refrigeration levels. If you have waste, low pressure steam available (as might be the case if you are using steam turbines - possibly with a back pressure of 30-25 psig), then you might make profitable use with a steam jet refrigeration unit. This unit is capable of doing two things: condensing waste, low pressure steam while pulling a partial vacuum over a water system and refrigerating it down to a maximum of 45 oF. This water is recirculated and only the vaporized portion is made up. If you have boilers on site, you may have ability to put such a unit together.

You may be located out in a desert and have scarce water capabilities; in that case, you have very few choices (or nil) as far as heat sinks are concerned.

I would not assemble your cycle as is shown on your flow diagram (which I suspect is nothing more than a simulator program print-out. For example, note that since you have a single-stage, totally enclosed cycle, there is no need to control the liquid refrigerant flow going to the evaporator. All the liquid you make in the Flash Drum goes there anyway; it has no where else to go. The same philosphy holds for the flash vapors and the evaporator vapors; these two vapor streams can easily be co-mingled in the Flash Drum and regulated into the suction of the compressor (through the KO pot). I see no merit in throttling a liquid feed to a refrigerant evaporator. The evaporator should be kept with all the tubes totally immersed in liquid Propane ( I assume the refrigerant is on the shell-side and that this is a flooded evaporator - the most efficient type and possibly a kettle type or similar.). I would normally place the Flash Drum in a location above the evaporator and feed the latter by gravity - with a semi-thermosyphon effect. This would eradicate most of the cited pressure drops in the liquid. I don't agree with the pressure drops accepted on the simulator's flow diagram; every pressure drop you build into a refrigeration cycle has to be paid for in energy demand at the compressor, sooner or later. This just represent wasted energy and gross inefficiency.

In the calculation worksheet, I left the evaporator heat duty as your initial basic data input so that you can try several capacities and see how this affects the results. I hope this product is of some help to you in analyzing your system. You have a very simple and direct refrigeration system, but one that is very very important to control because of the magnitude of the operating and maintenance costs involved. This is a very large unit and worthy of being well scrutinized and analyzed - such as you are doing. There are energy savings and economic rewards available to you in this unit, but you have to work at it in fine detail.

Lots of luck.

Art Montemayor

[gham]

Art Montemayor,

I really appreciate your effort and interest in helping me in optimizing our refrigeration cycle. The basic calculation you sent is very helpful and valuable. Thank you.

Before ending this valuable and fruitful discussion, I have couple of questions:

1- We have air-cooled heat exchangers (fin-fans) that condense the C3 vapor refrigerant from 187 F to 140 F before it goes to the receiver. Can we lower the outlet temperature (say from 140 to 120 or even lower) either by installing more fin-fans or modifying the pitch angle. In other words, would lowering the outlet temperature from the fin-fans help in subcooling?

2- Is there an optimum subcooling temperature? In other words, if we lower the temperature below 60 F, what is the trade off?

3- how does the steam jet refrigerant unit? would this contradict the fact that we are trying to reduce the steam consumption in the compressor turbine?

3- Is the following sentence correct: "reducing the flashed vapor in the flash drum reduces the overall refrigeration capacity"?

Comments:
1-As far as controlling the liquid in the flash drum, we have a control valve that controls the C3 liquid refrigerant flow coming from the flash drum and going to the evaporators based on the need of C2 Reflux. The more C2 reflux needed, the more C3 refrigerant going to the evaporators.

Thanks again

[Art Montemayor]

gham:

I believe I addressed most of these questions in the Excel Workbook I sent you, but I'll address them here nevertheless, in hopes of possibly obtaining comments and ideas from other forum readers and participants:

1) Any possible, additional cooling you can add to the Propane compressor discharge prior to reaching the Propane Receiver will mean less horsepower you have to consume to obtain the net refrigeration effect of your cycle. All you have to do is make a complete heat and mass balance of the overall system and you will see the effect that the subcooling has. This is why I set up the calculation worksheet the way I did for you so that you could see this net energy savings effect. As I have stated previously, it is always smart refrigeration engineering to make as much use of the natural occurring heat sinks around you, as low as you can get, prior to expanding the high pressure liquid into the Flash Drum. The more the subcooling of the HP liquid, the more the net refrigeration effect you create per lb of refrigerant circulated. The key here is the word circulated since that is the quantity that your prime energy consumer (the compressor) sees.

2) The optimum temperature that you can subcool down to is limited by the cost of doing this subcooling as compared to the savings in the energy that you realize at the compressor driver. This is not an easy optimization problem because it involves a variety and complexity of values that only you are in control of or have knowledge of: your energy costs, your cooling costs and availabilities, your capital money return constraints, etc.

3) I believe that I gave you a specific reference to finding out how a steam jet refrigeration system works in Ludwig's "Applied Process Design in Chemical and Petrochemical Plants", Volume III, Chapter 11. As I explained to you, this is not a contradiction when you have a source of available, low pressure, waste steam. This is even more applicable when you have a backpressure requirement on your turbines that forces you to generate LP steam and then have to condense it to return it back to your boilers. A steam jet refrigeration system can work with steam as low as 15 - 25 psig pressure. The amount of cooling available depends on the amount of steam & pressure available. All you are doing here is converting the useful, low level energy to a useful and profitable end simply because of the existance you have for relatively low level cooling. Normally, this is not the case; but in your application you could convert the enthalpy available in LP steam to profitable gain - if the LP steam is available.

4) I do not believe that your sentence, as stated and as it applies to your flow diagram, is a correct statement. The Propane vapor, flashed into the Flash Drum as a result of isenthalpic operation, has very little to do with the refrigeration effect taking place in the evaporator exchanger bundle. It is the amount of LP LIQUID Propane produced per lb of Feed Propane flashed that makes the difference in whether you get more refrigeration or not. The flashed vapor formed is related to the amount of liquid formed, but it itself is not directly related to what heat transfer occurs in the evaporator.

As you will note in my flow diagram I do not need a LP liquid throttling valve upstream of the evaporator tube bundle. For such an important and large application as yours I would always insist (at the preliminary engineering project level) on employing the most efficient type of evaporator known: a totally flooded evaporator with the refrigerant in the shell side. This could possibly be in the form of a kettle type. The reason for this is the obvious efficiency of the latent heat transfer taking place. This being an important decision made before all others, we wind up with a flooded evaporator that requires what it was designed to do: have a totally flooded shell side. A flooded evaporator cannot operate with its feed liquid being throttled. It must have full flow of feed 100% of the time in order to ensure that it is flooded. If you do not have a constant and steady heat load on a flooded evaporator, then you have to control the generation of LP refrigerant feed to it. This is conventionally done with a flooded evaporator - not throttling the feed to it.

Of course, you may not have a flooded evaporator design. This may be because of other factors - such as a lump-sum contract held by the plant's constructor who will always opt to "low-bid" on equipment costs. A flooded evaporator, although the most efficient, is more expensive because it is inherently larger physically - such as a kettle. With what little basic data I have on your application, this is all I can come up with. I hope the Excel workbook helps you to better understand what I have failed to explain to you in a clearer manner. Engineering graphics and calculations are far better than words to explain engineering principles and ideas.

Art Montemayor
Spring, TX

The plant I work at uses a very similar system as this. However we have a cooling water system that cools our compressor discharge to between 16 and 24 deg C. Previously we ran suction pressure to the refrig comps on auto and changed this setpoint based on refrigerant temp which varies due to ambient temperature affecting the cooling water. Due to upgraded control system we can now use a genlin algorithym to set this pressure based on propane to non-flooded condensor. The feeling is we can optimize power use by doing this. Can you advise me on optimum temperature and pressures or is this idea even valid?