13 replies to this topic

[Nasiruddin]

I am working on a refrigeration system design; my role is FEED design engineer. The objective of the refrigeration system is to cool the water which ultimately use for process cooling in the process plant.

Cooling water required at two temperature levels with following parameters:

Cooling water supply Level-1

1. Supply temperature to process plant: 20ºC
2. Return temperature from Process plant: 40ºC
3. Heat duty = 26 MW

Cooling water supply Level-2

1. Supply temperature to process plant: 12.5ºC
2. Return temperature from Process plant: 40ºC
3. Heat duty = 7.4 MW

The intended refrigerant is R-134a, Simplified Process flow diagrams are attached for the refrigerant as well as chilled water circuits.

My questions are following:

1. Role of economizer in Refrigeration circuit (as per my understanding as the name suggest to economize the overall refrigerant load by removing vapor after expansion valve).
2. Comment on my scheme the way I represented the economizer in Refrigeration unit. I used only flash economizer (no sub cooling)
3. Selection of refrigerant, the intended refrigerant is R-134a, but there is in-house discussion between Propane and R-134a, as R-134a has higher Global warming Potential (GWP) then Propane, phasing out of R-134a is also an issue. Estimated refrigerant flow (R-134a) based on above cooling loads and as per attached schemes is approximately 1000 Tonnes/hr, and estimated compression power is appox 11 MW.
4. Selection of refrigerant compressor type, intended compressor should be oil injected screw compressor.

 

Regards,

Nasiruddin

Attached Files

•  Refrigeration Circuit.pdf   106.07KB   82 downloads
•  Chilled Water Circuit.pdf   24.67KB   51 downloads

[shan]

The purpose of installation of economizer is to reduce power consumption of 1 stage compressors (K-8903A/B), which makes your refrigeration loop more "economical" (more cooling duty with less energy consumption comparing with single stage expansion).

[Art Montemayor]

Nasiruddin:

 

Shan is essentially correct, but the economy attained is not due to 2 stage expansion, but rather because the compression in the 2^nd stage is sub-cooled and approaches isothermal compression (but, of course, doesn’t reach it).  The term “economizer” is tossed about and used rather loosely in the refrigeration business – both commercially and industrially.  This is unfortunate, because the efforts made to try to reduce unit energy consumption are not understood as to concept and theory.  Consequently, there is a lot of misunderstanding that takes place all over the world with respect as to what an economizer is supposed to do and what it should look like.  As an example, I am attaching a workbook that shows an ammonia mechanical refrigeration system that I designed and built using a 2-stage system with an economizer.  Note that the ammonia economizer serves as a 1^st stage intercooler while dropping the saturated pressure of the ultimate liquid refrigerant to be ultimately expanded.  This concept is also what you are doing with your system – except that your flow diagram doesn’t clearly show that you are injecting the cold expansion vapors into the suction of the 2^nd stage of compression.  This 2-stage refrigeration system has very low unit energy consumption as compared if you did it in 1 stage.  You are also having to compress your refrigerant to the critical pressure in one stage before being able to condense it.  My system doesn’t have to do that since I use the economizer to cool the 1^st stage discharge vapors before they (and the expansion vapors) are introduced into the 2^nd stage.

 

Contrary to what you state, you are not “removing vapor after expansion valve”.  What you are doing is capitalizing on using a flash chamber that creates saturated liquid at the 2^nd stage compression suction pressure and thereby allowing useful recompression of the cold expansion vapors.  You are decreasing the amount of work required in the 2^nd stage.

 

If you are concerned about obsolescence or environmental problems, then I would select Ammonia as the refrigerant of choice.  Ammonia is far more efficient, economical, available, and environmentally sound than either R-134a or propane.  It is also safer than propane – a point I would emphasize.  There are no environmental concerns with ammonia.  I managed the design and fabrication of 2-stage ammonia refrigeration packages using Mycom screw compressors and these worked very well.  Ammonia refrigeration is not only the original type of refrigerant, it continues to be used world-wide, especially in the food processing business – which traditionally lack the quality and quantity of technical and engineering expertise that the chemical and petrochemical industry has.

 

Your flow diagram shows the symbol for a centrifugal compressor, not a screw compressor.

 Ammonia Refrigeration System Design.xlsx   85.83KB   97 downloads

[Nasiruddin]

Dear Art,

 

Thanks a lot for your detailed response and explanation regarding economizer function. I am now preparing a simulation model on the scheme I attached in my first post, and will try to compare all the refrigerants, i.e. R-134a, Propane and Ammonia. But I have one further question, NH3 I think should be used where low process temperatures required like -10 or -20 deg F. In my case I require two temperature levels i.e. 12.5 and 20 deg C.

 

However, I shall comeback once I finished all three refrigerant heat and material balance with my system.

 

Regarding flow diagram, you are right I shown the symbol of centrifugal compressor, because still it is not finalized to go with screw compressors.

 

Thanks a lot again for great help.

 

Nasiruddin

[shan]

Art,

Your heat & material balance table on the bottom of PFD is incorrect. When NH3 liquid Stream 12 is letdown from 12F, 40 psia to 15.5 psia, it becomes superheated vapor with temperature 6.2 F instead of saturated liquid -26F, 15.5 psia as you assumed. Therefore, the NH3 refrigeration loop is unable to provide the process demand cooling duty 182,890 Btu/Hr at -26F.

[Art Montemayor]

Shan:

 

Since this is the Industrial Professionals Forum, I have to assume that you are a practicing or graduate engineer.  Therefore, I have to respond clearly, concisely, and as one professional to another.

 

With all due respect, you are grossly wrong in whatever method(s) you have used to arrive at the conclusion that tells you that if you expand a saturated stream of liquid ammonia from 40 psia and 12 ^oF, you obtain a superheated vapor.  I strongly recommend that you seriously check your source of thermodynamic data that allowed you to arrive at such an erroneous answer.  I state this because:

• All the recognized and established thermodynamic databases that I use concur that my data and calculations are correct.  My data sources are NIST and “ASHRAE Thermodynamic Properties of Refrigerants”.  I have other numerous sources and they also concur with my data.  As a visual and basic example, I am attaching a Mollier Diagram for Ammonia and you (as well as everyone reading this) can easily apply basic, first-year thermodynamics and easily see that the isenthalpic expansion of the saturated liquid truly produces what my flow diagram and material balance show.  This proves your statements are wrong.
• Unbeknownst to you is the fact that the heat and material balance were done with a slide rule and are the official calculations I generated to design and build an actual refrigeration unit that recovered revert CO₂ gas from a tonnage dry ice facility that not only was built, but worked exactly as predicted and calculated for many years.  I don’t have to resort to the Mollier Diagram myself because the design worked – for years.

As a result of this discussion, I hope that one thing is very clear and well-understood amongst us:

 

The free, adiabatic expansion of a recognized saturated liquid refrigerant is carried out in an isenthalpic thermodynamic process and produces a 2-phase mixture of sub-cooled vapor and liquid – both of which are saturated.  This is so because – as can be seen in the Mollier Diagram, the expansion starts at the saturated liquid line on the LEFT HAND SIDE of the refrigerant’s saturation “Dome”.  Because of this location, the corresponding constant enthalpy line carries the expansion directly into the internal portion of the “Dome” – where you clearly have a saturated mixture of liquid and vapor.  You can easily calculate the corresponding amounts of both the liquid and vapor using the amount of the expanded liquid feed and the so-called “Lever Rule”.  Or, you can simply make a formal heat and material balance around the expansion valve .  I prefer the latter, using NIST derived enthalpy values.

 Ammonia PH Diagram.jpg   247.47KB   12 downloads

[thorium90]

I think the confusion arises frm where 13 was labelled on the diagram..

[Art Montemayor]

thorium90:

 

If Shan is confused with my sketch and heat/material balance, then that is my fault.  However, he/she clearly identifies what I noted on the diagram and the table:  saturated liquid ammonia at 40 psia and 12 ^oF (Stream 12) is expanded into an evaporator that is at 15.5 psia and -26 ^oF.   Both he/she and I agree on that being depicted.   What Shan stated was that that couldn't happen because the saturated liquid would produce a superheated vapor, and not a subcooled liquid + vapor mixture.   Am I not reading his last post correctly?   Perhaps I am wrong in understanding what was stated or in trying to communicate.  If so, please tell me where and how.

 

Steam 13 is the resultant, saturated vapor exiting the evaporator and its temperature and pressure is correctly identified by Shan as being predicted as -26 ^oF.   What "confusion" are you referring to?

[thorium90]

Ahh, sorry, my mistake. After abit more checking I realise the error I made. The stream after stream 12 is saturated liquid at ~15.5psia and -26F, and upon vaporization to saturated vapour takes up the 180000+Btu/hr.

 

To help shan:

(672.47-84.68)*310.9~183000Btu/hr

Attached Files

•  enthalpy.jpg   113.96KB   6 downloads

[shan]

Art,

 

Thank you.  I found my mistake in calculation of Stream 12 letdown.

 

As indicated in the attached flow diagram, I think it is better to insert a water cooler after the 1st stage compressor.  It will reduce about 10% of 2nd stage NH3 circulation rate and save about 10% 2nd stage compressor power.

 

Still slide rule in 1999?  I think I had my first calculator at least 20 years by then.  I remember I had a 486 with big black Hysim dongle hung on the back already.  Is that time when people worry about Y2K warms?  Please forgive me if my memory is not accurate.

Attached Files

•  Copy of Xl0000004a.xls   323KB   39 downloads

[paulhorth]

Shan,

So you still think you can teach Art Montemayor how to design a refrigeration system? ............

 

He didn't say he used his slide rule in 1999, either.

 

Paul

[Art Montemayor]

Shan:

 

And thank you for sharing your cycle modification.  I am very familiar with that variation of the cycle because I operated a couple of plants that way.  Although I have always been a hard proponent of using as much atmospheric cooling as possible at every compression stage within a mechanical refrigeration cycle, my calculation results showed that this concept was not as economical in investment capital and operating costs in spite of the power savings.  The reason I remember, was due to the relatively high Coefficient of Performance (COP) that my ammonia cycle produced.

 

When I added the capital costs for the intercooler exchanger, the piping and the incremental water tower investment together with the cooling water treatment and maintenance, the intercooler economizer came out ahead.  Of course, the application is sensitive to the type of available cooling water, its costs, the additional maintenance, and the cost of electrical power if an electric induction motor is the compressor driver.  The quantity and type of capacity unloading can also have an effect in comparing both types.  With an abundant and available 45 ^oF deep well cooling water source, I would certainly take a look at water-cooled intercoolers for this type of cycle.

 

The important point to make in this thread, in my opinion, is that the refrigeration system works as we have defined it herein.  It isn’t about who is right or who is wrong.  As long as we agree that we are discussing a viable, logical, and scientifically based application of a proven process we can only add to the value of the thread.  The refrigerant that can do the best job should be decided on logic, needs, and scope of work.  I suspect several compressors will have to be installed, from the power requirements identified.  What type of compressor will be determined by capacities, capacity load variations, local needs and restrictions, capital costs, energy costs, environmental concerns, and a host of other factors.  Today, because of environmental regulations and refrigerant prices, I see ammonia becoming more of a mandate than a selection – but again, that depends a lot on location.

 

I did my work exclusively with a slide rule for 14 years, until 1974 when I bought my first HP-45.

[shan]

paulhorth, on 14 Mar 2013 - 10:57 PM, said:

Shan,

So you still think you can teach Art Montemayor how to design a refrigeration system? ............

 

He didn't say he used his slide rule in 1999, either.

 

Paul

Paul,

 

Did I fail your qualification test that judges if anybody can teach Art Montemaryor how to design a refrigeration system?  Equity is the beauty and principle of cyber space.  Everybody is same whether you claim that you are a 106 years guru from graveyard or you pretend that you are a new born puppy from backyard.  All the same.  Right?  Therefore, you can teach me and I can teach you if I want to.

[gegio1960]

Just to note that these levels of refrigeration duties are usually managed by centrifugal compressors.

In fact, with screw compressors you probably will have to put more machines in parallel, due to the limited capacities.

Kind regards