Parallel VFD's

[The Plc Kid]

Load sharing is one thing, and requires comms between the drives. Separate sources though implies two completely independent drives that will both try to fight to control the same motor.

You would still need fast comms between the two, to keep their outputs exact, which implies one must be a master, which means it's not a redundant system at all, as something (probably bad) will happen if you lose one drive.

Input sync isn't an issue if you intend only to parallel the DC Bus link, but that is only part of what is being asked for here. Even for a computer data center, the usual practice is to simply have a duplicate of every HVAC plant where required for cooling, and only run one at a time.

Parker SSD - /shudder.

Correct it would be a master slave combo and to make it work both drives would have to sized to handle the full load and the software / firmware would have to be able to change master / slave orientation on the fly which i don't know if anyone does that. Example if you load the slave then the master would handle full load and if you load the master the slave would have to switch to master/ single drive mode and handle the full load. I would think that it is do able but i doubt anyone does this is there is very miniaml demand for something like this at best.

Again i don't know. Just kicking around idea's

 

[The Plc Kid]

Parker SSD - /shudder.

You got that right. They are a real PITA most of the time. These are systems that i inherited i would never use this setup myself. Most are on 800 HP motors running at 480 vac and thats the reason for a parallel system. A medium voltage drive should have been used here as we already use media voltage drives at this facility.

 

[rkicklighter]

Correct it would be a master slave combo and to make it work both drives would have to sized to handle the full load and the software / firmware would have to be able to change master / slave orientation on the fly which i don't know if anyone does that. Example if you load the slave then the master would handle full load and if you load the master the slave would have to switch to master/ single drive mode and handle the full load. I would think that it is do able but i doubt anyone does this is there is very miniaml demand for something like this at best.

Again i don't know. Just kicking around idea's

This is something similiar to what they're wanting. Each drive would handle the entire load and if input power was lost to one, the other could pick up the load instantly. Personally I don't think this is possible in the manner they want.

BTW, this may be a high-level datacenter.

 

[jrwb4gbm]

Just thinking outside the box. Perhaps you could use something similar to the drive-train that the Toyota Prius Hybrid automobile uses. This would suggest a electro-mechanical solution. It uses two motors mechanically connected to a single output shaft thru some type of gearbox.

 

[rkicklighter]

Just thinking outside the box. Perhaps you could use something similar to the drive-train that the Toyota Prius Hybrid automobile uses. This would suggest a electro-mechanical solution. It uses two motors mechanically connected to a single output shaft thru some type of gearbox.

Good thought, but I think these are axial fans with the fan mounted directly on the motor. Haven't seen the spec sheets on the fans yet.

 

[Christoff84]

Good thought, but I think these are axial fans with the fan mounted directly on the motor. Haven't seen the spec sheets on the fans yet.

Could you not have the 2 drives, 1 running and the 2nd standby. Power drops to the 1st, control redundancy kicks in and the 2nd one starts in flying start mode. Due to the fact the load will have a lot of inertia and spin down time when coasting, there should be no downtime on the fan. If you can prove this setup will only drop exhaust output by say 5% for 2 seconds it should be fine.

You could even go as far as having a small backup UPS to keep the 1st drive powered for a few seconds while the 2nd one takes over.

Don't forget 2 PLCs, 2 control relays per signal...etc etc etc etc...and double the bill too.

 

[JeffKiper]

and double the bill too.

At least double maybe triple.

 

[OkiePC]

Good thought, but I think these are axial fans with the fan mounted directly on the motor. Haven't seen the spec sheets on the fans yet.

If they are fans, I cannot understand why a momentary blip is not acceptable...

 

[OkiePC]

RANT: Is it just me, is is the emphasis on hurry up and do it cheap fully overtaken our industry in place of workmanship, craftsmanship and precision?

Is this what has led our fast-food culture to see redundancy as the cure for unreliable systems?

I still believe a properly engineered and installed drive is going to outlast a properly engineered and installed motor 90% of the time.

Why don't we focus on taking our time with the planning, spending big bucks once and doing it right with singularity?

To me, that's cleaner, more effective and just plain smarter than redundancy (with duly noted exceptions)

Redundancy in comms, yes, give it to me in quadruplicate if we can...sensors? Sure, throw some float switches at the end of the analog level sensors, I can use that sort of redundancy...

With PLCs...only if it's easy like controllogix.

But motors and drives? Motive devices in redundancy? I know there are good places where it fits and it works...

But redundant drives on a fan?

Isn't that wasteful, costly, complicated and unnecessary in many of the applications that come through this site?

[/rant]

 

[cwal61]

My first thought is: It's not the contractor's responsibility to design. That belong's to the Engineer that's going to stamp the print. The system being installed sounds like new work. Requiring Permits Inspection and Approval from the Local Authority. Therefore there should be prints to submit to the inspector's office. The Contractor is responsible for installation according to Engineer's Drawings and Local Code he is not allowed to change installation without Engineer's Approval. I would not touch this without approved plans and Engineered Prints. Let his butt fry not yours. The Project Scope sounds Ccmplex and you can not put enough money in upfront if something goes wrong in the end during start up. If not an Engineered Approved Drawing, Can you say T&M or Run like H-E-L-L

 

[rkicklighter]

Thanks for all the suggestions and comments. Yes, the scope on this is very complex and this particular customer always seems to make it 10x more complicated than it needs to be. We are the control system integrator and as such are responsible for the control architecture and design. That design revolves around specifications from the customer. On the parallel VFD idea and several others, we have pushed it back to the customer.

True, some redundancy is a good thing. Certain PLC's are redundant, ethernet and certain input devices. Like others, I don't see the need for redundant VFD's. To me it adds a layer of complexity that overrides any benefits.

They tend to over-engineer these sites to prevent failure. The sites are 24/7/365 but very few of them have any type of regular maintenance personnel who know anything about the control system.

Thanks Again!

 

[DexterIt]

One quick thouth that comes to my mind is,if they do work 24/7/365..........just like our site does,how do they manage in case of some ball bearings failures to the motors?what redundancy are they using for that failure?how much mainteinance personal do they have there?

 

[DickDV]

In my view, the redundancy engineering is pretty thin on the described system. There is all this concern over motor controls and none apparently over anything else.

Two posts above in this thread mention the motors. Here you have all of this expensive control redundancy and then you run the power down a single set of motor leads to a single motor thru a single coupling to a single fan pushing air thru a single ductwork system, etc,etc.

I would suggest that for solid electrical redundancy, they couple two motors end to end to feed the fan. Better would be the fan in the middle with a motor on each side with a common shaft with couplings thru the two motors and fan.

Feed both motors with independent VFD's powered from separate sources.

Provide matching control signals to each drive.

Power only one motor at a time. While unpowered, the motor simply freewheels with the other one and the fan. AC motors put very little load on the shaft when freewheeling like this.

Control timing, sequencing, e-stop, and probably other issues would have to sorted out but at least you have redundancy without the "weakest link" being in the electricals.

 

[rkicklighter]

Dexter, you bring up valid points that I have metioned also, but haven't got a response. I specifically asked, "Two supply fans, two exhaust fans per section with redundant everything control wise; what happens when you lose a fan blade or a set of bearings?" I got silence. True this is a very clean and climate controlled environment but nothing lasts forever.

As for maintenance personnel, I have been in many of these sites from different customers and all any of them have had is one permanantly assigned, 8 hour, 5 day week, contract person with others on call. Most of the time this person's knowledge is 98% HVAC with a little control thrown in and no PLC experience at all.

 

[rkicklighter]

Dick, I agree totally. This customer always goes overkill on the electrical and controls but seems to want the bare minumum on mechanical. It's their $$$$$ not mine.