Cooling tower and VSD

[Gauss2k]

Hello,

We are considering to install Variable Speed Drives on our 2 cooling towers to save energy and money. The towers's fan is 10 HP and the pump is a Monarch Industries BSEF-750 (Specs:

http://www.monarchindustries.com/v4/media/pdf/POP/bsef.pdf)

(7.5 HP).

Basically, the hot water arrives in a tank, enters the tower and is stored in another tank until it is needed by our machines. Presently, the pump and the tower fan are both activated at full speed when the output tank reach a predetermined low-level (to give an idea, it's activated about 5 min every 10-15 min). The output temperature varies, but since it's used to cool down molds, if it's hotter, we simply use more water. But this temperature variation requires more adjustments on the process and we would like to make it more constant.

I think we have a couple of options:

- Simply put a VSD on the fan motor to slow it down when necessary, and only activate it when the pump is working. (but this does not represent much energy and money savings)

- We put a VSD on the fan and the pump to slow both of them down. The pump only pumps the required water by our process (controlled by the tank level) and the fan regulate the output temperature.

I wanted to estimate the savings of that last option, but I'm not sure about how to do it with the pump. I can't find any chart on the manufacturer's site that tells me how the pump whould react at a lower speed (I also called them, and they told me they can't help me...). The only chart is at full speed (see the link I gave earlier). The tower is approximately 30 feets directly above the hot water tank.

Can anyone help me with these saving estimations? And do you think I should install VSD on both the pump and the fan or is it a waste of money in my case?

Thank you,
Stephane Richard

 

[jraef]

It appears to be a centrifugal pump (I am not a pump man actually), so the basic affinity laws should apply. That means that the power will vary by the cube of the speed. What will not be easy to predict without a lot more info is how much you can actually reduce the speed and still have usefull work being accomplished in your tower. For arguments sake however, lets say that you can run it at 70% speed constantly, as opposed to 100% speed for 30% of the time (5 min. every 15). At 70% speed, the power consumed would be .70 x .70 x .70, or roughly 34% of the relative power at full speed. You must however realize that you are only running full speed for 30% of the time so your savings may not necessarilly come from direct energy reduction. You will however have much less wear and tear on your motor doing it this way. This is a gross over simplification of course, but you should be able to see where it is heading.

There are some excellent software programs that will calculate the energy savings for you, but you will need to first know what speed you can effectively run the pump at, and to get that, you probably have to buy the VFD! I would try to sell your bosses on the reduction of wear and tear, then find out if you saved energy after the fact and go for a bonus!
Check out

http://www.drivesurvey.com/index_library.cfm?feature_category2=Energy Savings

Click on the Power Tools tab as well and they have a free software program for determining energy savings.

"Venditori de oleum-vipera non vigere excordis populi"

 

[quark]

The primary inefficiency in your system comes from the storage of cooling water. Your cooling tower water outlet temperature will be lower than the ambient temperature(depending upon the area WBT and Approach of the cooling tower) and you are heating it back by storing. You will get better results as long as you run the cooling tower during your process.

Jraef has explained clearly about the probable savings by using VFDs. Danfoss has developed a software program which calculates the process savings that exactly match your case and fortunately it is free if you build good relation with them.

My suggestion would be to have constant speed fan and run the pump with variable speed. (your savings will come from the hidden advantage that, air resistance to the fan reduces at reduced water flow rate and the power consumption goes down). Switch on and off the fan as you do it now with respect to the cooling water outlet temperature. Check the temperature rise of the cooling water just as you receive it and also as you start using it. If the difference is morethan 0.5[sup]0[/sup]C, you can seriously think about changing your procedure.

Good Luck,


Eng-Tips.com : Solving your problems before you get them.

 

[jherbert]

I think you have opportunities to reduce
energy, a few thoughts:

- as quark stated, your cooling towers (CT)
provide a variable cooling water temperature
dependant on the external conditions. Of course,
these vary over the course of a year.

- The system has 4-5(?) starts per hour,
VFD's will reduce wear and hence maintenance.

Perhaps re-engineering would offer savings too.

- The link provided for the pump(s) is
self-priming type. Are the pump(s)
installed above the water level?
(Self priming pumps expensive, would be
pretty usual in this standard application).

- Say 20 minutes at full load per hour
satisfies your load, this implies the installed
capacity exceeds your present needs.
Installing a VFD will certainly save energy.
Reducing system capacity will improve
efficiency, lowering operating costs and
perhaps, requiring a smaller VFD's

- Improved efficiency, and perhaps cold
tank insulation could reduce cooling
water volume and increasing the moulder
output per hour.

cheers,

 

[Gauss2k]

Thank you for your answers,

Here's a few more details. The hot water temperature is usually around 165F (74C) and the "cold" water temperature is about 85F (29.5C). So the "cold" water is still hotter than ambiant temperature. I don't think that storing it for a few minutes causes any significant problem.

Quark: I agree that changes in the air resistance will change the power consumption, but does a change in the water flow really have a big impact? The simple fact of turning at full speed, with or without water to cool, requires a current near the rated current of the motor, am I wrong?

And concerning the change in the water flow, I have thought about another possible problem that would maybe explain why they installed these over-sized motors on the pumps. If we reduce the water flow, especially during very cold winter days, the water may freeze in the tower. I'm trying to get more informations about the "minimum required flow" (if it is possible to know it) from the company who sold us these towers. If our flow is near that minimum, I guess the best I can do is to add a soft starter on the fan to at least reduce the stress on the mecanical parts. (and forget about a bonus for the energy saving!!)

Finally, to answer jherbert: the hot water is brought to an underground tank (right under the floor level) by gravity. Since the pump is on the floor, it must be self-priming to work properly (well, as far as I know! My pump knowledge is quite limited)

Stephane

 

[GaDawg]

From my experience with a considerable amount of variable speed pumping systems for HVAC and other industries, it is interesting now to see CT fan control follow suit.
There is a lot of potential for energy savings here, but don't let oversimplification fool you. For the pump, The affinity laws will work great up to a certain point. For instance, assuming the pumps are located below the CT supply and discharge tanks, you will always have a certain amount of head pressure to overcome due to elevation and any required pressure to fill the supply tank. Unlike the closed loop scenario you would experience with a chilled water system, the speed of the pump will follow the system curve only to the point where this minimum pressure is reached. This might mean only a 25% speed reduction even at 75% to 0% (shutoff) flow. Where as with true affinity law characteristics, at 50% flow, the pump would be at 50% speed delivering an 87.5% reduction in BHp. However, on a system that might continually run, but not at full demand, even a 5% reduction in speed is significant.
Your greatest potential for relative energy reduction might come from your fans in this case. Assuming the CT is sized for say a 2% WB design temp., most of the year the fan capacity is oversized. This could also be highly variable throughout the day from nearly no airflow required to full airflow required with equivalent static head differential. Therefore, you potentially could make full use of the affinity law characteristics. Ultimately, you will want to discuss the variable speed application with all you equipment vendors, since there might be underlying issues that come into play, such as resonance and motor dv/dt problems. I hope this helps.

 

[patrickjmp]

don't vary the flow across tower without checking with the tower manufacturer. If you go below the minimum flow on the tower you will dry out the fill and damage it. This will in turn destroy tower performance. Also varying the flow across the tower will change the outlet temperature which would in turn would vary the speed of the tower motor. To save energy, vary the tower fan speed based on the outlet temperature leaving the tower before it enter the tank. Any decent VFP with PID control should handle this very well.

 

[Gauss2k]

After a check with the manufacturer, the actual flow was, like I thought, used to prevent water freezing in winter. So we'll not change the pump speed and instead just add a soft starter on the fan. (A VFD on the fan doesn't generate enough ecenomy in our case)

 

[cub3bead]

Have you considered switching to two speed fan motors instead of VFD?