PID Loop for Fan Control

[comeng 0]

Need some help with setting up a large fan on PID control, supplier of control system SLC 5/03 has configured PID loop in his way with a large acceleration/deceleration setting (approx 60secs) within the AC inverter (255kW) which seems reasonable given the size of the fan however the drive manufacturer says that normally when using PID control they set these figures around 3 seconds and control speed up from start via the PID loop, control supplier says does not think this can be done with AB PID statement as effectively using two different ramp figures one for startup and another for normal running conditions. Fan currently very unstable when setup this way All suggestions gratefully received Phil

[BobLfoot 302]

What it sounds like you are describing you need is to use the SWM {Software Manual Mode} for your ramp up and ramp down and use your PID for "stable state" adjustments. My idea would work itself out like this. 1. Assume 50 hz is the speed at which you wish PID control to begin. 2. Assume 15hz is the minimum running speed for your fan and motor. 3. Startup Sequennce is as follows: A. Place PID in SWM Mode. B. Issue 15 Hz comamnd to inverter and start a delay timer of 1.5 seconds C. When Delay Timer is done add 1 hz to the inverter comamnd value, reset and restart the delay timer. D. Repeat Step C until the Inverter HZ command reaches 50hz, then start a holding delay timer of 15 seconds. E. Once Holding Timer is done Switch to PID auto mode and let PID control Fan. 4. Ramp Down to Stop is the inverse function.

[paulengr 44]

Easily done. Depends on what you are trying to achieve on startup. You can run switched PID's...implement 2 PID loops, or you can store the gains in a separate set of floating point values and switch them on the fly. An example of the 2 PID loop problem is when you have two different goals. For instance if you have a system where you normally control air flow but since it's a combustion process, you don't want to generate CO either and you have an oxygen sensor. So one PID loop could be set with an oxygen set point of 1% and another control loop could be set based on CFM or pressure. The output is selected based on whether O2 is less than 0.5% (oxygen loop controls) or not (CFM loop controls). However, I'm guessing that your startup problem is that you need to come up slowly and that the integral gain is causing windup and oscillation. I had a similar problem recently on a 30 MMBTU/hour burner. The gas is controlled by a cross-connected pressure control valve so that the gas ratios to the air. The air side is controlled by a PID loop under PLC control. During startup when everything is cold, if I just let the PID loop operate at normal speed, it would go to 100% fire as quickly as possible. Since the flame isn't very stable when everything is cold, it would blow the flame out. So I used the output clamping values. I set up a timer on startup set for 100 seconds. The timer .ACC value is written to the PID loop's output high value clamp. This way on startup the PID loop slowly ramps from 1 to 100% in 100 seconds. Once it reaches full output, the timer is done and the PID loop is allowed to operate normally. An outside vendor did it a slightly different way. Since it's a burner and on startup virtually all burners simply ramp up to high fire, they just set it into manual control and ran through a programmed ramp before transferring to automatic. Usually the reason for setting the acceleration/deceleration values in the drive for a fan application under PID control is that you don't want to allow the drive to ramp up/down so fast that the regeneration circuit can't absorb all the energy from the fan inertia. Depending on how you set your gains, you can achieve anything you want. By the way, make sure that you have the anti-windup bit set, too. If you don't, then you will experience windup and get all kinds of wild things going on especially on startup....in fact, this could be your problem.

[comeng 0]

Thanks to Bob and Paul for two interesting suggestions I am going to look into both when I can get back onto the plant and play in safety Problem stemmed from long period of running drive in manual where the long acc/dec times in the drive wee essential to protect the fan, the argument arises that we need PID control but occasionally will need manualand will therefore need to protect the drive, we might have to have switched values within the drive Incidentally I have thought of incrementing the SP in the PID loop but don't know if this has any advantages/disadvantages in comparison to the other suggestions Will get back to you when I have the answer Phil