21 replies to this topic

[shekhar dhuri]

I have following situation for the LPG Export Pumps.

LPG from Debutaniser Column is at 14 barg @ 65 0C. This LPG is very close to its bubble point. The LPG will be stored in the horizontal mounded bullet and will be pumped using export pumps. Conditions of the export pumps are : Flow = 8 m3/h and differential head = 1200 m. NPSHR as per the suppliers of the LPG export pumps is = 2 m (minimum). The possible options to provide higher NPSHA are :

Option 1 : Elevate the LPG bullet
Option 2 : Use trim cooler upstream of the LPG Bullet so that LPG in the bullet will be at sub-cooled conditions.
Option 3 : Use special service LPG pumps with low NPSHR, low flow and high head.

I am thinking of going ahead with options 2. Since increasing elevation of the storage bullet or using special pumps have implications on cost and schedule.

Can you please confirm my understanding of putting Trim cooler.

Regards,
Shekhar

[fallah]

 

Shekhar,

 

I would prefer the option 3 because:

 

Option 1 cannaot be a proper option otherwise the needed value the bullet has to be elevated would be low

 

Option 2 also cannot be a proper option because adding trim cooler not only might compensate the higher fixed cost of the special LPG pump but also will result in higher operating cost than LPG pump option. Indeed, dependence to trim cooler would decrease the reliability and availability of the pumping system.

[paulhorth]

Shekhar,

 

I have worked on a similar project a few years ago.

 

Personally I would prefer Option 2, the trim cooler. You should be able to get to 50 C or perhaps 45 C with an aircooler in almost any part of the world. The heat duty will be small, leading to a small low-cost unit, though I cannot be sure if it will be lower cost than a special low-NPSH pump. The fan could be turned off for large parts of the year when the air is cold enough, reducing running cost.

 

Alternatively, you could run the debutaniser with overhead temperature at 50 C, thus avoiding the trim cooler, though this will mean a lower pressure in the column, which might mean that a pump would be required for the bottom product to go to storage instead of runnig down under pressure. That might still be the simplest option.

 

Low-NPSH pumps for LPG (C3-C4 mixture) could be vertical canned pumps with the impellers below ground level, which gains a meter or more of NPSH,  though this raises problems of maintenance and casing drainage. However, you may need to provide sun shielding over the pump suction piping from the storage. I worked on a project where the solar heat gain was identified late and the sun shields had to be fitted just before startup.

 

It would be expensive to raise the storage bullet by a sufficient height because of the need to cover with a mound. By the way, this mound is a vitally important protection against fire - so it's best to leave the storage as it is.

 

Paul

[S.AHMAD]

Shekhar

1. I agree with Fallah comments that option 2 is not a good permanent solution.

2. Indeed, adding cooler will reduce the temperature and hence reducing the vapor pressure. That increases the NPSH_a.

3. However, you must also consider that the cooler is subjected to fouling. Fouling will eventually increases the temperature and pressure drop. This combination of increasing pressure drop and increasing LPG outlet temperature would further reduces the NPSH_a. It could be even worst than existing problem.

4. Have you look into increasing the inlet piping size? and remove the inlet strainer if any?

5. Reducing the pump frequency would also help in reducing the NPSHr. Changing to double suction would also help.

Edited by S.AHMAD, 23 December 2013 - 02:37 PM.

[Bobby Strain]

Since you will need a multi-stage pump, use a vertical canned pump as Paul suggested. You can dig the hole as deep as you like. You won't get a chance to do it right later.

 

Bobby

[ankur2061]

Shekhar,

 

One of the important point that has been missed out is whether the LPG bullet is an existing one or a new installation is being discussed.

 

If an existing installation is being discussed the recommendations Paul has given are worthwhile exploring including the low NPSHr vertical canned pumps. BTW, for a cooler in such a clean service such as LPG, fouling should be the least of the concerns.

 

However, if you are discussing a new installation the most worthwhile solution both from the viewpoint of economics and complexity of the system considering special canned pumps and / or trim cooler would be to have the LPG bullet installed at an elevation which would give you enough margin between the NPSHa and the NPSHr for a conventional API 610 horizontal mounted centrifugal pump.

 

Regards,

Ankur. 

[S.AHMAD]

Shekhar

1. Sorry that I did not get the right picture of your plan to install a cooler up- stream of bullet and sub-cooled the LPG. This plan will never work out the way you think. The bullet storage have normally vapor space inside, am I correct?. If so then the bullet pressure shall be the vapor pressure at whatever temperature that you call it sub-cooled. With this scenario your NPSH_a remains the same as before. You are wasting money, your cavitation problem remains the same.

2. The right approach is to cool the LPG upstream of the pump downstream of the bullet storage. This will reduce the temperature and hence the vapor pressure at the pump inlet also be reduced. Therefore, NPSH_a increases.

3. As Ankur pointed out, LPG cooling fouling is not so much expected, probably this is better option.

 

Lavi,

Yes you have good ideas. But remember, for LPG there is no such thing as storage at sub-cooled temperature. The storage is always at its vapor pressure at the storage temperature.

[fallah]

 

2. The right approach is to cool the LPG upstream of the pump downstream of the bullet storage. This will reduce the temperature and hence the vapor pressure at the pump inlet also be reduced. Therefore, NPSH_a increases.

 

AHMAD,

 

The point appears to be theoretically attractive!

 

Considering the suction line to be as short as possible, an increase in NPSHA due to cooling effects might be compromised with a decrease in fluid pressure due to passing through a cooler; result: no increase in NPSHA

Edited by fallah, 25 December 2013 - 03:51 AM.

[S.AHMAD]

Fallah,

I agree, the system must be properly designed to ensure that the NPSA gains due to temperature reduction must be greater than the pressure drop across the cooler. Otherwise the reverse may be true.

[paulhorth]

Ahmad,

 

 

But remember, for LPG there is no such thing as storage at sub-cooled temperature. The storage is always at its vapor pressure at the storage temperature.

You are of course absolutely right, and I was therefore completely incorrect to propose subcooling in my post No. 3. I didn't think carefully and missed the point you have now made, forgetting what I had known on an earlier project. 

I am embarrassed, as I should be.

 

With humble regards

Paul

[Lavi]

"But remember, for LPG there is no such thing as storage at sub-cooled temperature. The storage is always at its vapor pressure at the storage temperature."

 

 

@ Ahmad Sir:

 

Agreed. And couldn't agree to your statement any better. I had reverse approach to solve the problem. It was futile. Assuring you, this shall never happen again.

 

@Paul:

Looks like you are not the only one who is embarrassed.

 

Despite this, I would still propose having an Air Cooler at the upstream of my storage tank. I have two reasons for this.

 

1. First one is obvious. It is going to increase the NPSHa of my Export Pump. By using an air cooler I would be bringing down my temperature from Point A to Point B (Refer attached diagram) thereby increasing the density of the fluid in my sphere. This in turn would increase my pump's NPSHa.

 

2. Cooling the fluid before storage, is going to reduce my storage pressure. As we know, storage pressure would be the same as vapor pressure at that sub-cooled temperature. This would mean I am decreasing the upper limit of my operating pressure range. And hence, would be cutting down the thickness of the storage vessel (by few mm?)

 

Please correct my understanding. I am still in the process of learning.

 

Attached Files

•  untitled.bmp   1.12MB   20 downloads

[fallah]

Lavi,

 

The points might theoretically be attractive!

 

1) A few degrees decreasing of LPG temperature--->A few percent increasing of LPG density---->A few percent incresing of NPSHA; Is it worth to consider an air cooler for such low increase in NPSHA?

 

2) It is the side effect of the temperature decrease and hasn't direct relation to NPSHA but the effect on design pressure of the storage vessel cannot be significant. Indeed, if a few degrees decreasing in LPG temperature can lower the vessel thickness it might lower the minimum design temperature of the vessel such that may lead to vessel material change!

Edited by fallah, 27 December 2013 - 05:45 AM.

[Raghaghi]

Hello for all members,

I'm happy to joint this site now.

I have some questions as below about LPG pumping that I would thank you to help me by your reply:

1) Do you have experience of cavitation in the LPG pumps? I designed and manufactured one multistage centrifugal pump to pumping LPG at +40C, and suction pressure equal to 11 barg. NPSHA was about 3.0 meters, but as NPSHR can be higher than NPSHA beyond the rated point, or in other word, NPSHR is not less than NPSHA along whole the operation rang, I'm worry about cavitation at inside of the pump. I want to know that is there need to ask our client to install one another small OH2 pump before main pump to feed required suction head at all the working conditions?

2) bubble implosion in the pump because of cavitation, is explosive and dangerous, or not? Is it similar to the other liquids, such as water and will results just pressure drop at suction, or No? I'm worry about explosion at site because of cavitation inside the pump.  

3) I selected the pump component's material in a similar way to other applications from API 610(selected material class is S1). So, is there any body to know that, should I consider any specific material for at least rotating parts or no? Because as I know, stainless steels rotating parts are sparking if those touching each other and for this reason, these material is not applicable for explosive pumping liquids.

Your reply to my questions above, can help me strongly and I'm thank you for your kind reply.

Best Regards,

 

[Lavi]

@ Fallah Sir,

You seem to be a man with a lot of common sense!

 

1. Agreed to your first point. I prepared a calulation sheet for cases with and without an Air Cooler to confirm the same. For the benefit of young readers like me, sharing the output. To my amazement, I am finding NPSHa without an Air Cooler to be greater than NPSHa with an Air Cooler. This is contradicting to the theory I had assumed. It is because the pressure drop across the air cooler is so high, that the increase in density is not capable of compensating the loss.

 

Please refer the attachement. Also note that the data obtained is from a running plant which my team had designed in my country, Tunisia.

 

2. Agreed to your second point also.

 

@ Raghaghi,

 

1. This whole thread is about the concern you mentioned in point 1. I have a limited experience of designing LPG plants to 2 projects. And in both these projects, we had made use of OH booster pumps to feed the export pump. However, both my designs have an OH3 as suggested by vendor and not OH2.

 

2. Nothing is explosive until it fills the three corners of the FIRE TRIANGLE. You shouldn't be worried about this because your Operational & Commissioning team shall purge your storage tank with Nitrogen first and then with Fuel Gas if required. Nitrogen should bring down your oxygen % to less than 2% or 3% (depending on purity of nitrogen you use). And then bone drying using Fuel Gas should ensure that your oxygen level is lesser than 0.5%. This shouldn't be dangerous unless you have oxygen ingress.

 

3. I believe such exotic materials shall be well advised by the Vendor. Or may be you must wait for Seniors to comment. Ahmad Sir has an exhaustive experience in LPG plants.

Attached Files

•  NPSHa Cal.xlsx   13.06KB   28 downloads

Edited by Lavi, 28 December 2013 - 03:43 AM.

[S.AHMAD]

Paul, Lavi

 

We all make mistake since nobody is perfect. What important is that we learn from our mistakes.

[Raghaghi]

Dear Lavi,

I'm thank you very much for your comprehensive reply.

but I will be glad if I get the others' help in this explained issues above.

Dear Lavi, I will try to upload a movie about LPG plant explosion later.

Best Regards,

Raghaghi

[shekhar dhuri]

Hi,

Based on process data & unavailability of the canned Pumps I am going ahead with the following ;

• Trim Cooler will be installed downstream of the LPG Bullet in the pump suction line.
• Trim Cooler will be cool the temperature of LPG from 60 degC to 50 degC. LPG composition Propane=49%, i-butane=15%,n-butane=26%,i-pentane=5%,n-pentane=5%)
• With 10 degC cooling, vapour pressure of the LPG has reduced by 2.5 bar (from 13.5 bar to 11 bar)
• Due to 2.5 bar reduction in vapour pressure, NPSHA has increased drastically. Pressure loss due to installing air cooler is insignificant as compared to the increase the vapour pressure.
• As informed by Mechanical Engineer working on my project, presently, there is no benefit related thickness of LPG bullets.

Thanks all for your comprehensive inputs.

Regards,
Shekhar

[Raghaghi]

Dear Lavi,

May be you already saw this movie, but I just share it for you to know my concern.

https://www.youtube....h?v=HN-DShSmo_g

It seems that it should be LPG plant, because of Fire Wall generation after explosion.

Best Regards,

Raghaghi

[ankur2061]

Shekhar,

 

The attachment provided is a very good guide for "Liquefied Gas Transfer Equipment". It also provides a great explanation for the boiling phenomena for "Liquefied Gases". I would recommend it for anyone who is dealing with transfer of "Liquefied Gases".

 

Regards,

Ankur.

Attached Files

•  Corken_Liquified_Gas_Transfer_Equipment.pdf   7.21MB   47 downloads

[Yajnavalkya]

Dear All

 

We had tried out costing of both the options, to reduce the LPG temperature after the bullet my air coolers against having vertical canned pumps for a PSU in India.

 

Storage was in mounded bullets and maximum LPG throughput rate for Tanktruck Gantries, Wagon filling and Carousal Fillingwas 300 m³/hr.

 

Installing canned pumps worked out to be cheaper in our case. FYI, site was a coatal location in Maharashtra, India with maximum and minimum ambient temperatures being 45 °C and 12 °C.

 

I am just curious as to the site location for the specified problem, because, in India, LPG bullets design pressure as per codal requirements is of 14.5 kg/cm²(g) at 55 °C. However, cylinders had 16.9 kg/cm² (g) at 65 °C. Hence the bullets are not designed to hold LPG @ 65 °C.

 

I am also attaching a snapshot of ASTM D1835 which specifies the max. Vapor pressure of Commercial Butane, Propane and Commercial PB Mixtures for general info.

 

 

 

Attached Files

•  Vap Press_ASTM D 1835.bmp   5.49MB   18 downloads

[Art Montemayor]

I'm sad to read that some our members still have problems understanding what basically constitutes a saturated fluid system.  I have been pumping LPG since I graduated from college in 1960.  I never have experienced any pumping failures - in either both American continents, Africa, the Middle East, and Europe.  I also have read the Corken Pump paper (since I have used a lot of Corken pumps), but I have issue with their use of the term "boiling" within the LPG system.  I firmly believe there can be no boiling of liquid in a saturated storage LPG tank.  All that happens when there is external heat applied is that the system adjusts to a higher vapor pressure - in accordance with its vapor pressure curve.  There continues to exist pumpable liquid within the storage vessel.  And because liquid continues  to exist in the vessel, it is totally pumpable by the user.  What some people make mistakes about is providing a natural vapor "trap" at a suction line high point.  This does not provoke "boiling"; all this is is an extension of the storage conditions and, as such, is a location that is prone to contain a saturated vapor (which causes a pumping problem).  But this is not boiling.  It is merely a saturated liquid seeking to reach equilibrium with its counterpart vapor.

 

If you allow free, sloped flow for saturated liquid to enter the eye of any pump under a proper Net Positive Suction Head that is superior to the pumps minimum NPSH, then the pump will pump without any failure - regardless of the liquid's temperature (within reasonable ambient limits).   I have not only pumped LPG (one of the easiest fluids I have pumped), but have even pumped saturated liquid Oxygen and Nitrogen at cryogenic temperatures (lower than -250 ^oF) using a positive displacement plunger pump (which is more difficult than a normal centrifugal) in ONE STROKE to a pressure of 3,000 psig (suction being at 10 psig).   And I did this CONTINUOUSLY, 24/7 (24 hours a day, 7 days a week) - without stopping for at least 2 months at a time.  Now, that is tough.

 

That is why I believe that pumping LPG is a piece of cake and is being debated as if it presents problems, when in reality there is no problem here except one of not understanding the simple basics of fluid conditions and fluid flow.

[S.AHMAD]

Dear Art

Thanks for your good comments. I have came across several severe cavitation problems of pumping saturated liquid in some existing operating plants. It is sad to say that some engineering companies sometime assigned these jobs to inexperienced process engineers who do not fully understand hydraulic systems. The company should assigned experienced engineer to supervise the young and inexperienced engineers as part of learning/training of the young engineer by doing the real job. It is a good training exposure but they need experienced mentor to ensure good and effective design.

Edited by S.AHMAD, 03 January 2014 - 09:00 PM.