Fan outlet pressure vs temperature 2

[cokeguy]

Our glass container manufacturing plant has several 350HP VFD-controlled fans for bottle cooling. They normally run close to full speed (60Hz), since the outlet pressure setpoint is normally set to 60"H2O, the fan´s rated pressure at rated airflow. However, since we would like to have the highest possible pressure WITH the lowest possible temperature, we need to know if there are temperature v.s. airflow v.s. pressure fan curves, or formulas or rules of thumb, which could tell us how much the outlet air temperature rises with respect to pressure and airflow in centrifugal fans.
We want to see if under some cases the overall cooling effect could be increased by LOWERING the fan´s speed, and thus the temperature, right now we only control outlet pressure and disregard temperature.
Related comment: sometimes when we stop the bottle machine and airflow goes to a minimum, the outlet pressure actually decreases rather than increase, an effect which has been puzzling us a little, since normally you expect the pressure to rise in a centrifugal fan when you demand less airflow under constant speed. Anyone out there can point to some literature on this subject? Thanks a lot for your help.... Normally the manufacturer supplies pressure vs airflow vs power vs system curves, but no curves or formulas take into account outlet temperature rise. Thanks again for any comments on these matter.

 

[MintJulep]

From:

http://www.learningincontext.com/PiC-Web/pdfs/AirSys.pdf

Temperature rise = Total Pressure Rise /(2.7 * Fan Efficiency)

Temperature rise in degrees F

Total pressure rise in inches of water.

 

[EnergyProfessional]

what you are experiencing might be surge region of the fan. Google for it or read this

 

[Drazen]

that looks like upside-down-engineering to me.

i am wandering why you are so certain that pressure should be maximized? it is air flow rate that brings you cooling effect - assuming that we talk about heat rejection; pressure only serves to transport that flow.

you can google "fan curve", "ventilation system curve", "ventilation system working point" to get basic idea of it.

 

[cokeguy]

I understand your point Drazen, we use fan speed to control outlet pressure (we don't have outlet flow measurement) and thus airflow, which makes the actual cooling job. But we were wondering if perhaps there was a point where the temperature increase due to the increased pressure would actually offset the increased airflow cooling effect. However, from MintJulep´s post we can see that the actual temperature rise due to increased pressure is negigible in our case. I'm not sure though, and would certainly appreciate a comment, if that relationship mentioned by MintJulep holds in general throughout the whole fan´s operating range.... thanks for your comments.

 

[macmet]

Can you please explain the process a little better. I have never seen a bottle cooling process. Are you blowing the air over the bottle, into the bottle, both....?

 

[cokeguy]

In a glas bottle machine, airflow is directed from the fan's outlet plenum into pass-through openings in the cast iron bottle molds. The machine can have up to 10 molds working at the same time, and each mold has its own automatic damper that opens only when the corresponding mold is actually in operation (each mold can operate independent from the others). Therefore, when the whole machine or at least several molds stop working their dampers remain closed, and that´s when we see the effect of pressure actually decreasing while airflow decreases.
For example yesterday, the machine worked with only one or two molds for 8 hours, airflow obviously was minimal, and pressure dropped from 60"wg to 58"wg, causing the automatic pressure control to increase fan speed from 56Hz to 60Hz, and as a result temperature rose from 26 degC to 46 degC!!!
Obviously the formula [Temperature rise = Total Pressure Rise /(2.7 * Fan Efficiency)] didn't hold in this case, because pressure remained the same but air still heated up, with airflow reduced to a minimum. Is this because this formula applies only under constant speed or under non-extreme conditions like the one I'm mentioning? Is there another formula that approximates fan behaviour under these extreme circumstances? Thanks for your comments.

 

[macmet]

Is this a new set-up or did you have it working before?

Do you measure the pressure upstream of the manifold (in the plenum)?

 

[cokeguy]

It has always been that way, on every machine, in fact this setup up is standard in all glass bottle manufacturing plants, but only on that machine have we noticed that behaviour of "surging" (or whatever it is) causing that pressure decrease/temperature increase when airflow decreases.

 

[macmet]

Do you see pressures that you expect when all 10 dampers are open? What happens if all 10 are closed?

You've checked that the pt is piped properly with no blockages?

 

[cokeguy]

Yep, if all or most of the dampers are open, pressure and temperature behave as expected. If none or only one damper is open, pressure begins to drop and temperature rises sharply.

 

[Drazen]

your new details indeed give clearer picture.

from what you described i would firstly assume that there is flaw in cooling-blowing concept IF the mold dampers do not posses pressure-independent flow control ability. if those are ordinary on-off dampers, system curve "jumps" over fan curve and maintaining constant differential pressure by changing fan rotation gives different fan-system interaction for each and every on-off combination.

that means fan should be "ordered" different pressure for every combinations of damper open (all open, all but one open..., etc.), but i am not sure can you set it up at fan speed controller.

in that case primary solution could be disconnecting all mold dampers leaving them open all the time.

i am not sure how would it affect energy consumption. if in you current practice energy consumption would be significantly increased, than setting up control solution as mentioned could be long-term cure.

 

[macmet]

If all the other machines run with the temp and pressure conditions you would expect why would this one act differently thermodynamically? Have you tried running with the fan in hand and monitoring temperature and pressure?

I still do not understand your process completely, but if there are other machines that work there is no reason for pressure vs temperature to change.

 

[cokeguy]

thanks again for your comments. In the glass container manufacturing industry, maintaining stable mold cooling regardless of the day/night temperature swings has always been an issue, but the truth is that I´ve never seen a serious study or modelling of the effect of temperature swings on the bottle or mold cooling process. Some plants/lines use a temperature-compensated pressure control, where for example for each 2.7 degrees of temperature increase, pressure setpoint increases 1 "wg, to (supposedly) maintain a consistent cooling effect.

In our plant, operations personnel blame the air temperature swings for some of the bottle quality problems, so that is why they want us to maintain a stable air temperature, which obviously is not feasible cost-wise. And that is why we have been looking closely at those temperature increases related to airflow, trying to figure them out. I'll keep you posted, while I try to understand a bit more those temperature effects, since I am just the control/instrumentation guy resonsible for the control systems, with very limited HVAC knowledge.

 

[cokeguy]

Follow-up:
I made a test today on one of the fans, at fixed speed (56 Hz), checking the plenum pressure under various conditions:

The whole machine (10 molds) running: 58"H2O
machine stopped (0 molds): 54"H2O
1 mold running: 56"H2O
2 molds running: 58"H2O
3 molds running: 59"H2O
4 or more molds: 58" H2O

so as HerrKaLeun said, it is definitely a surge condition, right?
I had seen surging in compressors, but not on centrifugal fans. I suppose it only happens with higher outlet pressure fans, as the link posted by HerrKaLeun shows fan curve examples with a surge region, but the pressure never actually decreases at low airflows, it remains more or les stable, but outlet pressure is only 10"H2O.

 

[willard3]

What type of fan is this?

Only axials and FC centrifugals operate have surge regions in the fan curve.

 

[cokeguy]

It is an FC centrifugal fan.

 

[willard3]

Fan surge, then, is a completely reasonable explqanation.

 

[willard3]

You can get a velomneter and plot the fan volume or you can get the fan curve from the original manufacturer. Plot your volumes and pressures on the fan curve.......is the measured pressure/flow in the surge region of the fan?

 

[BronYrAur]

What is the manufacturer and model number of the fan?