VFD Cooling 2

[tulum]

Hello folks,

Just a quick question regarding VFD cooling. We need to install a 150HP drive in a 80"H x 40"W x 24"D nema 4x metal enclosure. Because of the enclosure-drive size ratio we need some form of cooling. The powers at be do not want any sort of liquid cooling system, and ventalation in a 4x enclosure becomes difficult.

Any suggestions? Estimating the VFD losses at 4500W, and the ambinet outside the enclosure at 35degC, I am at a loss.

From the above the surface Area = 84.4ft^2 therefore the heat load is 4500/84.4 = 53W/ft^2?? Seems pretty High!

Any help appreciated.
TULUM

 

[DanDel]

Check out this link for Vortex cooling:

http://www.process-controls.com/Pelmar/Vortex_Enclosure_Coolers.html

 

[tulum]

Oopps... forgot to mention... compressed air is also out... don't ask me...

Any other hints??

 

[Mike2468]

If you can not install fan ventilation and you can not use vortex cooling, I assume that you already rejected using a cabinet air conditioner. One remaining option is to make the entire cabinet into a heat sink. Use an aluminum cabinet and attach large external heat sinks to it. We did a PVC factory where the air tight cabinets had a bank of heat sinks on the outside back wall, from top to bottom. External heat sinks will get dirty so plan on using 2 - 4 times as much surface area than you think that you need.

Put a circulation fan inside the sealed cabinet so the VFD's heat gets spread out over the entire inside area.

And your external ambient temperature is going to exceed 35C on some days, so plan on that too.

 

[cuky2000]

Check if this ifo could help.

http://www.reliance.com/pdf/drives/whitepapers/D7752.pdf

 

[jraef]

Thermoelectric (Peltier effect) cooling systems can be made to NEMA 4X specifications. They will be expensive for this size application, but they may be your only alternative. If you are not familiar with them, they use solid state cooling plates to move heat from one side to the other without moving parts like a compressor. Since they are essentiall a heat sink on each side, one side can be passed through a sealed box with no air exchange. The only caviat is that the external fins will be aluminum, maybe anodized, but still unsuitable for long-term exposure to certain types of corrosives. Here is an industry forum link that gives a list of providers.

http://www.zts.com/modules.php?op=modload&name=Web_Links&file=index

"Venditori de oleum-vipera non vigere excordis populi"

 

[ko99]

TECs are great but I don't think they make sense for your application. They are very inefficient. You'll need to add a ~4kW power supply and the enclosure will have to dissipate the TEC heat (4kW) plus the VFD heat (4.5kW).

Consider conducting the heat from the VFD to the inside of the enclosure with thick aluminum or copper brackets, then add a large external heat sink to the outside of the enclosure. Ideally the brackets would connect directly to the VFD surface at one end and the sink at the other end (the sink is part of the enclosure wall and the perimeter would need to be sealed). Thermal pads or braids may be required for vibration isolation.

ko (

http://www.ecooling.biz)

 

[ScottyUK]

Although not big enough for this specific application, being limited to 37kW, Allen Bradley have a flange-mount variant of the Powerflex70 which is designed to mount the heatsink through an aperture in the enclosure so that the drive body is inside but the heatsink is outside. The drive comes complete with a seal to IP66. We are expecting to use some of these units before long - if anyone has any field experience of these units I would appreciate the feedback.

You could investigate "heat pipes" as a means of moving large amounts of heat energy from an internally-mounted drive to an external heatsink. I expect that you will need to do some surgery to the drive in order to mount them, but they are capable of shifting quite amazing power from a hot semiconductor to a remote heatsink.

Google "heat pipe" for loads of hits.

Good explanantion at...

http://www.dynatron-corp.com/products/heatpipe/heatpipe.asp?page=heatpipe

Products at...

http://www.aavidthermalloy.com/products/heatpipe/index.shtml

----------------------------------

If we learn from our mistakes,
I'm getting a great education!

 

[jraef]

ko99,
I appreciate your input on TECs. Having never used one yet I had no idea it would take that much power to operate!

The problem with your suggestion is in the basic design of a standard VFD. It already will have a air-exchange system in its design to dissipate into the enclosure, but my experience is that getting the heat from the inside of the box to the outside is damned near impossible without some sort of active system. Even heat exchangers cannot typically keep up with the gain, it has to be either a compressor, TEC or vortex tube if available.

The problem I found with vortex tubes is the air volume necessary will translate to a lot of additional air compressor HP, making them the most expensive to operate of any choice. I did one project where it was necessary (in the days before TECs), but for a 48 x 36 x 12 box with just a PLC inside it translated to 7.5HP of additional compressor load!

BTW tulum,
Using the Hoffman air conditioner calculation program for your application, it comes up with a minimum of 1600BTUs of cooling, will consume only 400W, but is at best NEMA 3R, even if you order the expensive stainless steel version.

Really the best solution might be ScottyUKs suggestion of finding a VFD in that size which allows mounting of the heat sinks on the outside. You may need to relinquish the N4X criteria however, because most of them are not rated for that, especially if they need fans on the outside heat sink assembly. IP66 is close, but does not meet the corrosion resistance specs for N4X.

"Venditori de oleum-vipera non vigere excordis populi"

 

[ko99]

Thanks, jraef, I'm not very familiar with VFDs.

Also, I wasn't very clear: the bracket(s) should conduct heat directly to the inside of the enclosure WALL, not simply to the air inside the enclosure. And ideally the wall is a heat sink with fins external to the encloure.

It's along the same lines as ScottyUKs suggestion, except his external heat sink is part of the purchased VFD, whereas my brackets may require some 'surgery to the drive'. As Scotty also suggested, the brackets could be heat pipes rather than alum or cu. There are some tradeoffs there.






ko (

http://www.ecooling.biz)

 

[electricuwe]

To me the solution of flange mounting like described by ScottyUK looks very promising and as far as I remember there are also other drive manufacturerers providing similar solutions.

In the past I have designed several other power electronics systems that way. The main drawback is that most people want to mount enclosures side by side with the back against the wall and you can not mount a 150 kW VFD in the door as you can do with a heat exchanger or a cooling unit.

 

[telasin13]

This is more a question than solution I think. Would it be possible to mount plate type evaporators to the exterior walls of the enclosure, insulate them to improve efficiency and prevent condensation. Use a hi temperature refrigerant like mp39 so as not to bring the interior walls below dew point.
This would in a sense make the enclosure an inside out refrigerator. The surface area should provide adequate heat exchange .
Now don’t jump on me .I’m in the learning process myself here.