3 motors common duct 2

[pcubed]

I have a customer who has 3 100hp 460vac motors that blow air into one common duct run. The system is controlled by maintaing static pressure of .65"WC in a 6' round duct. The motors are controlled by three different VFDs. What is the most cost effective way to run these motors? I believe that running each motor at a lower speed will save more energy vs running 1 or 2 at higher speeds. I need proof to present to my customer before I can make any changes to the program.

Pcubed

 

[VxA]

My impulse would be to say that running one motor at full speed would be cheaper than running three at a lower speed. It probably depends on the characteristics of the VFD though.

Don't have any proof though, just a gut feel.

Thanks,

VxA

 

[CJCPE]

I agree with VxA. The efficiency of a VFD and motor combination is highest at maximum speed and torque. Providing the flow with 1 or 2 motors will be more efficient.

The reason is that some of the losses are constant. If the VFD is energizied but not running, it still consumes some power. If a motor is running, but the shaft is not coupled to anything, it still consumes some power. You may be able to find curves or data on some manufacturer's web site to confirm this. Many manufacturer's have programs for estimating energy savings and some of them may provide sufficient information to confirm this.

 

[ozmosis]

pcubed
what are the 3 x VFD's doing at present? are they all running together on the one common pressure feedback?

What has your customer got to lose by trying the various options out? It is usually the other way around where you have to convince somebody of a payback by actually fitting VFD's, but in this case, you already have them and it is a case now of optimising the performance.

 

[pcubed]

Currently 2 of the VFD's are running at 70% to handle the load. However this will increase as demand for cooling increases. All 3 use the same pressure sensor and setpoint. My goal is too come up with the most efficient way to run the system and get the customer to purchase the re-programming based on an attractive ROI.

 

[jraef]

You need more information.

If it is a true Variable Torque load, the affinity laws apply, so power consumption will vary by the cube of the speed. For instance a single motor running at half speed will consume 1/8th of the power of that same motor running at full speed. But for multiple motor installation comparisons like this you need to also know what amount of flow you get at reduced speed from each motor, it's most likely far less than speed reduction percentage. How much less will make all the difference.

Let's compare the scenario to having one motor running full speed, full air flow, vs all 3 motors running at reduced speed / flow. For simplicity, efficiency issues are being left out on purpose for now because although it makes a difference, it's only slight.

Best case; assuming that you might get 1/3 flow at 1/2 speed, then 3 x 100HP motors at 1/8 power is 37.5HP to get the equivalent flow of 1 motor running full speed at 100HP, a theoretical savings of 62.5HP.

A worse case scenario is that to achieve 1/3 flow, you must run the motors at 70% speed. So the power will vary by .7 x .7 x .7 or 34.3%. So now your 3 100HP motors running at 34.3% power to get the equivalent flow just became 103HP, i.e. not a great plan.

So as you can see, you MUST know the speed - flow data from the fan system before determining this equation.

Eng-Tips: Help for your job, not for your homework Read faq731-376

 

[BJC]

Hopefully you don't have a leaky damper on the third fan as was discussed in thread 181-27567.

This seems like it might be more complicated than fan speed vs pressure. What kind of fans are they using? You may want to look at the fan curves. Power for centrifugal fans gos up exponetially with fan speed.
Each running fan will see a .5" WC pressure differential but how much power each consumes to provide 1/3. 2/3 and 3/3 of the flow for differnt total flows I think can be determined for each fan.
Sorry
I may be way off on this but it sounds like a fun problem. You may want to find the mechanical group for fans where the smarts on that subject are found.

 

[DickDV]

I would expect that running these fans like a sequential control system on pumps would work out to be most efficient.

The first motor runs on the VFD up to full speed. If more is needed, it throws the second motor on at full speed and backs its variable speed back to minimum speed. As demand continues to rise, the variable speed motor again gets to full speed. If more is required, the third motor is thrown across the line and the variable speed motor backs down to minimum speed. As demand continues to rise, the variable speed motor again increases up to full speed, at which point the system is max'ed out.

A further imbellishment of this is to add rotation to the pumps so periodically the first online motor becomes the variable speed unit, the second online motor becomes the first online unit, and the variable speed motor becomes the second online unit. That way, the service hours is balanced between the units and wear and life cycle can be expected to be more nearly equal.

I don't see any good reason why this can be done on fans with or without the rotation feature.

 

[waross]

If the fans have backflow dampers then the comparison os reasonable.
If the fans must be running fast enough to prevent backflow the economics change.
We haven't yet mentioned friction drag on the fan blades.
By the time you get through trying to construct efficiency curves for the VFDs at various speeds and loadings, it may be easier and more accurate to connect a wattmeter and measure the various combinations of speed. Can you "waste" air to vary the loading on the system so as to get comparisons at various load levels?
respectfully

 

[LionelHutz]

As others have said, you need more data.

I do like DickDV's idea but I would vary the speed of all of the running fans, not just one. The turn-off point for one fan when, say for example, all three are running would be a little below the airflow you get when two run at full speed.

 

[pcubed]

I just got some more data from the field - There are isolation dampers for each unit so that when the fan is off the damper closes. Currently 2 fans are running at 60% to maintain the .65" static. My technician performed a test by running all three fans at once. 2 of the fans ran at 42% and one ran at 40%. The fan data for all three is

59500 CFM
970 RPM
100 HP
78.66 Breaking HP
460Vac

Those numbers are for full speed 60HZ.

I may have to take waross's advice and put wattmeters on each fan.

Thanks to everyone for all the help so far.

 

[jraef]

Not being a mechanical guy, I'd guess that you have some difference in flow restriction that is accounting for that difference in percentage, 2 of the fans have more restriction that the 3rd, probably additional bends etc..

Assuming that you mean percentage of speed, you most likely have this scenario:

2 fans at 60% speed = 200HP * .216 = 43.2HP total

2 fans at 42% speed = 200HP * .074 = 14.8HP
1 fan at 40% speed = 100HP * .064 = 6.4HP
21.2HP total

You will have more losses in the 3 motor scenario, but not as much as the additional savings. For argument's sake, let's say your throughput losses are 5% at 60% speed, and 9% at 40% speed (these are somewhat typical loss values). Input power draw then becomes:

2 fans @ 60% speed = 43.2HP / .95 = 45.5HP
3 fans 2 40% speed = 21.2HP / .91 = 23.3HP

Higher loss percentages, but still far less power consumed.

You will have more maintenance required by running all 3 motors continuously as opposed to alternating them to even out the wear and tear, but that is somewhat offset by the reduced loading on the motors. Overall, you still have the ability to accept the higher power consumption when servicing one motor, but you get to take advantage of the lower overall power draw on a continuous basis.

By the way, check to see if your VFDs have a kW output reading on the display, or at least the ability to send the kW value to the analog output as a 0-10VDC signal. Most of them do. Then all you would need is a meter to read the voltage on that signal to get the kW.

Eng-Tips: Help for your job, not for your homework Read faq731-376

 

[sreid]

I've just got to ask. Exclude back dampers. If paralled fans are running at different speeds, isn't there likely to simply be back flow through the slower speed fan? Air moves from high pressure to low pressure.

 

[jraef]

Again, not a mechanical guy, but my understanding from working with them is that the same flow restrictions that require the other 2 fans to run faster would make the static pressure at the point of common coupling to be the same.

 

[pcubed]

Air does move from high to low and it does it by following the least resistive path. Because all fans are running at pretty much equal speeds they are each creating the same amount of pressure in their respective duct. When the ducts join the duct area is larger resulting in less static pressure at that point and for quite a ways down stream. With all of the supply drops in to different spaces along the way the lowest pressure is always down stream.