Jump to content


Torque control on fans


  • Please log in to reply
7 replies to this topic

#1 Guest__*

Guest__*
  • Guests

Posted 26 September 2005 - 06:28 AM

Today I received a trade magazine that had an article in it telling how Scheider Electric had replaced some old soft starters with their new ATS-48

In this article they say "With nameplate currents of around 430 amps they were drawing some 500 percent at start and taking 15 seconds to wind up to speed. The soft starters that were in place were simply time voltage ramp. This led to large transients and harmonics being introduced to the network"

Later they go o to say " By installing an ATS-48 on each motor for the now six fans, Schneider was able to limit the current to 280 percent and get the fans running efficiently within 10-12 seconds"

I don't know all the ins and outs of soft starter design but is the torque control that speak of in the ATS-48 that good that it could drop the current and time by this much over an older designed soft starter? I thought that the amount of current was dictated by the load.

Also they intimate that the harmonics problems were cured too? Can you do this in new soft starter design?

Just some observations. Any comments?

#2 marke

marke

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 2,604 posts
  • Gender:Male
  • Location:Christchurch, New Zealand

Posted 26 September 2005 - 08:56 AM

Hello Anonymous
Register on the forum and you can get email notification of replies. - dont hide under a bush!!

A solid state soft starter employing SCRs to control the voltage applied to the motor can only provide two means of voltage control to the motor.
Option one is standard phase control of the SCRs and a derivative of this is a form of limited cyclocontrol. The voltage control can be incorporated within a feedback loop or an open loop control algorithm to alter the starting characteristics.

If we look at the operation of the motor under a constant frequency environment, the start torque is reduced by the voltage reduction squared.
If the slip is higher than 10%, there will be a very high current at full voltage (Typically 600%-800% on older motors and 700% - 850% on modern high efficiency motors). In order to accelerate the motor and load to full speed, there must be sufficient energy transferred to the load. This energy is essentially proportional to the product of torque and time. Higher torque requires a shorter time for a given load inertia.
If the frequency is altered, the slip can be reduced and provided that the slip is less than 10%, the start current for a given torque output will also be reduced. Cyclocontrol techniques are a means of providing a subharmoninc modulation of the supply frequency. This will provide a low frequency component in addition to the line frequency component. An 18 SCR cycloconverter provides a good means of "stepped" frequency control at low frequencies relative to the supply frequency. This is often used for propulsion control on large vessels. The soft starter uses only six SCRs and so the level of control is much restricted and the frequency steps are quite large.
This technology is used by some soft starter manufacturers to provide a low speed jog forward and reverse. Because the current is both phase control and cycle stealing, there are often major supply issues related to this form of control. Additionally, the "Pulse currents" can cause additional stresses in the motor.

By using a subharmonic generated using cyclocontrol techniques, it is possible to provide a higher initial torque at zero speed that can be used to provide 'break away' torque to get the load to begin to rotate. While the average current will be lower for the torque developed, the peak current will still be high with a very low frequency modulation. Up to a relatively low speed, the subharmonics can be stepped to provide acceleration, but the majority of the acceleration period has a very high slip component and the performance would be the same as classic phase control line frequency technology.

There is no doubt that an open loop voltage ramp start will result in a much higher start current than other closed loop starting methods. When the current is higher, the torque is higher and the acceleration rate is also higher. This will result in a shorter starting time, not a longer start time. One of the potential problems with the voltage ramp start can be the initial start voltage. If the initial start voltage is too low, there will be an extended period of very low acceleration and this could extend the start time. A low start voltage setting will also result in increased rotor heating.

Closed loop starting techniques enable feedback to modify the start voltage profile in a manner to give a smoother start. The torque feedback model can reduce the start current as the motor approaches full speed, but for equal acceleration at any value of slip, the start current will be the same, irrespecitve of the feedback model used.

The prime advantage of the torque start is that it mor accurately controls the torque during start. If we compare this with the constant current soft start, as the motor approaches full speed, the torque increases quite rapidly. This can cause issues such as belt slip. The torque start model will reduce the start current as the motor approaches full speed and eliminate the acceleration rush to full speed. This will extend the start time for the same maximum start current, but give a more linear acceleration.

Bottom line.

  • Cyclocontrol techniques can provide a boost of torque to enhance breakaway but do not enhance the total start curve.
  • Voltage ramp starts will result in the highest start current.
  • Torque starts provide enhanced acceleration curves but longer start times than constant current starters with the same maximum start current.
  • Increased start current --> reduced start time.
  • Reduced start current --> increased start time.
  • Reduced SCR conduction angle --> reduced voltage, current and torque.
  • Reduced SCR conduction angle --> increased harmonic current component.

      I have difficulty with the concept of reduced start current and reduced start time if the previous starters were correctly set up. Voltage ramp start with artificially low start voltage, or based on SCR Diode technology could unrealistically extend the start time with elevated start currents, but this would be due to inadequate setup and so the comparison is not representative.
      If cyclocontrol technology is used, the transients and harmonics on the supply are much more severe than standard phase control. This can be comfirmed by many users of starters with cyclocontrol.
      Assuming standard phase control of SCRs, a lower start current is due to a greater voltage reduction. This increases the harmonic content of the current. A high start current has a much lower percentage of harmonics than a low start current for a given motor at an equal amount of slip.


      ;p;Marketing liscence??;p;

      Of course we must remember that the articles in that magazine are written by and paid for by the "advertiser" so are not necessarily as impartial as we may expect.

#3 msdaif

msdaif

    Intermediate Member

  • Full Member
  • PipPipPip
  • 27 posts

Posted 26 September 2005 - 12:42 PM

The Altistart 48 (ATS-48) is a product of Telemecanique which belongs to the Schneider group. It is a controller with 6 thyristors for the soft starting and soft stopping of 3-phase squirrel cage asynchronous motors.

The data sheet says that the nominal motor current (In) is adjustable between 0.4 and 1.3 times the current rating of the starter. Maximum starting current is an adjustment of 1.5 to 7 of In with limitation of 5 times the current rating of the starter.
Telemecanique says that the torque control of the Altistart 48 applies the torque to the motor during the entire starting phase.
The accelerating torque can be virtually constant over the entire speed range. It is possible to set the ATS-48 to obtain:
1) a high torque on starting for a rapid motor speed rise whilst limiting its temperature rise,
2) a lower accelerating torque at the end of starting for gradual loading.
Applications: centrifugal pumps or machines with high resistive torque on starting.

Telemecanique claims they have applied their patented technology:
Torque control System (TCS). No more details.

However, I am not sure if their technology would be different from what we know of softt starters

#4 marke

marke

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 2,604 posts
  • Gender:Male
  • Location:Christchurch, New Zealand

Posted 26 September 2005 - 06:52 PM

Hello msdaif

If we assume from the published information that the ATS-48 uses classic soft starter control with phase controlled SCRs, and the closed loop algorithm based on the absorbed KW, then there is no way that there shoud be a reduced starting current and reduced starting time, unless the "reference" is set far from optimum with a long time at low voltage (and low current) resulting in little or no acceleration torque.

Best regards,

#5 jraef

jraef

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 683 posts
  • Gender:Male
  • Location:USA, California

Posted 26 September 2005 - 09:41 PM

Just an opinion, but I would say that most likely they were replacing some really old analog soft starter that was 3 SCR-3 Diode design, and the end user had used his own clamp-on meter or a simplistic separate digital meter to read the original starting current, and had no idea how to interpret their readings. The original claim of "500% current for 15 seconds" sounds a bit far fetched to me for a fan started with a reasonable soft starter of any sort. Most likely the RMS current was much lower at that time, but they had no means of reading it properly. After the Schneider starters were installed, the current was taken from the Soft Starter display which would have displayed the corrected RMS current as measured by the soft starter itself.

If it were not a measurement error, they would be violating the laws of physics to say that they accelerated faster with less input energy. It just can't happen. Without knowing the voltage of the system we can't determine power values involved, but one thing never changes: it takes a fixed amount of power to accelerated a load from a dead stop to full speed. Although there is no such thing, let's make up a unit of energy called Amp-Seconds since that is all we have to work with. 500% current for 15 seconds represented 32,250 Amp-Seconds of energy on that motor. 280% current for 12 seconds represents 14,448 Amp-Seconds, significantly less power regardless of the voltages involved. There is no way that you can accelerate a load with less energy, unless in the first instance a significant amount of energy was being wasted as heat somewhere, and that would have to be in the motor. If there was that much energy being absorbed by the motor in the first instance, it would have been overloading in my opinion.

I agree, "marketing license".
"He's not dead, he's just pinin' for the fjords!"

#6 niallnz

niallnz

    Intermediate Member

  • Full Member
  • PipPipPip
  • 27 posts
  • Location:Christchurch
  • Interests:Electrical Engineer, working with industrial controls, motors and drives etc

Posted 27 September 2005 - 01:34 AM

The soft starters that were replaced were PDL Flowstart or Flowtronic or somethin like that. They have 6 SCR's and were set up for voltage ramp with current limit.

The fans are axial and are inline with each other. When I had an oportunity to work on these fans there were only two in each line. They are contra-rotating and had delayed starts to reduce the peak current. The first fan has some load on it as the train going through the tunnel draws a lot of air through the fan, but the second fan is esentially starting on load and the motor is turning backwards at the time of starting.

At the time of them first going in they failed alot until the poor PDL engineers figured out that were starting under a significant load and put in really beefed up SCR's and then set them on a current limit that got the load away in a reasonable time, 15 sec. the excess start time limit was 30sec, I think.

Now that the trains going through the tunnel are longer they have had to extend the scheme to six fans, what they haven't said is that I think they also beefed up the power supply and can push the whole thing a bit harder.

#7 Guest__*

Guest__*
  • Guests

Posted 30 September 2005 - 03:11 AM

Thanks for your comments.

I know that the ATS-48 doesn't have an "inch" speed function so they can't be using cyclocontrol techniques.

I still find it hard to believe that their patented torque control can start the fans with less current in a shorter time than the old timed voltage ramp PDL units. I could understand one or the other separately but not both together.

I like jraefs comments. They make sense.

Considering that Schneider now own PDL, it makes me laugh how they are dissing the old PDL units which are really their own product!!

#8 jraef

jraef

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 683 posts
  • Gender:Male
  • Location:USA, California

Posted 30 September 2005 - 04:26 AM

I believe that Schneider forced PDL to dump their soft starter product line prior to their taking over. I would think that would probably be to avoid bad press. Don't know that for sure, but it is coincidental that the two events happened in close proximity to each other.

I used to work for Square-D here in the US right at the time of the takeover by Schneider. We were told to dump the OmegaPack soft starter line (among other things) with no replacement in the wings, just quit selling them. About 2 months later the anouncement of the "merger" was made, and immediately we had to sell the Telemecanique ATS23 soft starters. The Frenchies really love that thing, and they will tolerate no other roosters in the henhouse so to speak.
"He's not dead, he's just pinin' for the fjords!"




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users