9 replies to this topic

[simadri]

Dear Groupmembers,
I have a reactor operating at 110 deg. C and atmospheric pressure. The source of heat is LP steam (@ 50 psig). Hence limpet coil has been provided to the reactor for steam supply. Requirement of steam is 260 kg/hr to the reactor as per heat balance. Inlet steam line size to the limpet coil is 40NB. The reactor is vertical equipment with flat roof and torispherical bottom. The capacity of the reactor is 6.5 m3. ID and TL to TL length of the reactor are 1.9m and 2.2m respectively. Liquid shall be filled inside the reactor upto 2.1m height from bottom TL. I have to calculate the following.

• Size of limpet coil
• Pitch of limpet coil
• Heat transfer area available in the reactor (since the complete shell is not covered with limpet coil, I am confused how to calculated heat transfer area available).

[ankur2061]

simadri,

Several references that you can look up for design of limpet coil also known as half-pipe coil jacket.

http://www.cheresour...d-vessel-design

http://www.cheresour...cketed-reactor/

http://www.cheresour...half-pipe-coil/

http://www.thermoped...om/content/547/

Hope this helps.

Regards,
Ankur.

[simadri]

Ankur

I have gone through the references. But I did not get answer to my specific questions. I got only one input related to my question.
In a thread, Art mentioned that heat transfer area available for reactor with limpet coil is 60-65% of total shell area.But did somebody calculate the heat transfer area?

[Shivshankar]

Simadri,

Hope this helps


http://chemjunction....limpetcoil.html

Regards
Shivshankar

[Art Montemayor]

Simandri:

I don’t know who recommended a limpet (half-pipe) jacket on your reactor – or for what reason – but the application is headed for a lot of problems – economic and practical.

In my opinion and experience, you are going to initially run into some bad and costly fabrication problems:

• You are only transferring 520, 000 btu/h, so simple and readily available heating panels are applicable to the total surface of the reactor;
• Since the reactor is rated for atmospheric pressure, the required shell will be relatively thin. Unless the shell is at least 3/8” thick, the concentrated heat build-up during the welding of the half pipes will be such that the shell will warp and deform. This is a very labor intensive and technically slow procedure that is costly.
• The forming of the “limpet” coil is also labor intensive and costly, together with its cutting into half pipe coil. This is not the case when using less expensive steam panels.

You will also have practical design and operating problems:

• You will have to design your “atmospheric” reactor for the external pressure in the jacket; this means you will require a vacuum mechanical design – which will call for a thicker than usual reactor shell – and with possible shell reinforcement to withstand the external steam pressure; This is not the case when using steam panels.
• You will get a very large pressure drop for the condensing steam in the coil due to the excessive large distance that the steam has to travel within the coil; you can fabricate parallel coils, but this will complicate the coil welding and installation and make it even more costly;
• You will find that as the pressure in the coil decreases due to control and the condensation taking place; this increases the specific volume of the steam and increases the velocity within the coil, causing erosion and wear and excessive pressure drop; this requires you to increase the condensing pressure or reduce the steam flow rate.
• The total available surface area to the coil is much less than if you employed steam panels directly on the external shell surface.

From the above, I hope you can see where this is all headed. I am reciting the above from personal experience in operating plants in the past. I hope this experience helps you out.

[Chris Haslego]

Art is right on the money here.

I don't think there is anyone left who can fabricate these types of jackets at a reasonable costs.

Clamp on dimple coils to the side of the tank and bottom (with cut-outs as necessary).

Flash weld studs for coil mounting, apply heat transfer mastic to maximize heat transfer between the coils and the tank shell, insulate, and I'm afraid that you'll get what you get in terms of heat transfer. Luckily, you don't need much.

These types of clamp on jackets are O.K. for maintaining a temperature, but if you need quick ramp up in temperature, that likely won't happen. If you have a viscous liquid in the tank....it's going to be really slow.

[simadri]

Art,
Before opening the website today morning, I was expecting your post. Hence I am very happy because your posts are always informative and there is something new to learn from your personal experience.

At first, I had considered plain jacket. MOC of the reactor is Titanium which is so costly. As you have mentioned, shell will be thicker to sustain the pressure of steam (design pressure 6 kg/cm2(g)). After calculation, the shell thickness is 18mm. Since this is uneconomical, I thought of three different options.

• Internal circulating steam coils

Since the reactor contents will come in contact with the internal coils, their MOC must be titanium. Hence I have rejected the option.

• External limpet coils

I have not much knowledge on limpet coils. But the shell thickness has been reduced to 5-6mm using limpet coil. Hence I thought of the option.

• SS reactor with Ti cladding

From my point view, this is a good option. But I do not know what shall be minimum thickness of Ti cladding.

I have heard about steam panels for this purpose first time from you. But I know that steam panels are used to heat buildings for fire prevention. If possible, please provide me some details/drawings on steam panels so that I can suggest the same to client.

[Art Montemayor]

Simadri:

Since you hadn't mentioned it, I had no idea that your material of construction was titanium. The importance and the impetus to use external, independent panels increases.

Attached is the kind of information you are requesting. This type of panel is made by Mueller or Tranter. I have used Tranter, and can recommend them since they did the job(s) that I specified for various applications. I remember using 30 psig steam on one application.

As Chris has remarked, the usual method is to apply a thermal mastic under the platecoils to increase the heat transfer rate via conduction. This system works, but as I have preached for a long time now in our Forums, vessel jackets have no role in acting as process preheaters or steady state process heaters. They simply do not respond to the basic scope of any reactor. If your client requires a rapid heat up or cool down, then he/she would be well advised to apply a forced circulation, external heater and pump. If that is too expensive, then an agitator and a tank internal coil would be the next option. There is no way to avoid the inherent trade offs when dealing with heat transfer. If you want efficiency, economy, and speed - then you have to PAY.

I hope this information and the experience helps you out.

Attached Files

•  Tranter-Platecoil Applications.pdf   881.41KB   1740 downloads
•  Tranter Platecoil Information.pdf   580.24KB   1005 downloads
•  Platecoil Engineering Shapes Brochure.pdf   1.54MB   1059 downloads
•  Mueller Heating Panels TP-412-4.pdf   169.44KB   1624 downloads

[simadri]

Thank you Chris and Art for sharing valuable informations.

[T.S.MURALI]

You can’t achieve your objective with HTA, space, pressure drop, and heat absorption/heat load constraints.

You cannot provide the HTA in a constrained space.
You have to achieve by forced circulation of liquid/fluid using appropriate heat exchanger.