VFD's and Cabling

IGBT drives : Assuming that your application is an AC drive, standard practice is to provide " inverter duty " motors as per IEEE; the xlp cable may be ok but again the best industry practice is a cable similar to "drive rx" cable , in metal conduit or with metal armour, which has the three phase conductors in a tight wrap with three ground conductors ; and finally, at 200ft ,you may need dv/dt filters at the motor terminals ( usually available from a good quality VFD supplier) . hope this helps to get started - check with Baldor also.

Thanks. I have checked with baldor and one was a rep and the other was a baldor motor repair shop. Unfortunately I got two different answers. One said you could operate the motor at 4:1 and the other said 10:1. For our cooling tower operation we are set to operate at 3:1 or 2:1. I think this would put us in safe territory. As for the cable. the runs are all under 200' and 80% is in conduit. The two longer runs at 180'. I plan on using dv/dt filters on these. The cable tray run, even though easier concerns me. I will run it in conduit.

What about feeds to the drives. I have encountered engineers who are running armored cable to feed the drives. Is this necessary?

I would prefer to use drive rx cable, but the cost would exceed our roa.

Thanks again.

You don't have to have drive rx cable to feed the drive; it is necessary on the load side of drive to motor. Regular cable in metal conduit ( standard) is fine for primary side from supply to drive. How many HP is this ? you might have to consider harmonic currents on the supply side of the drive if the HP is> 10% of supply( i.e. 100HP : 1000KVA transfo ); if so ,then add a 3 or 5 % choke on input to drive, but again check with manuf ( if you don't get good engineering help from Baldor, then check with the manuf suggested by the other commentator( good advice).

Greetings,

As there are different types of IGBT drives on the market with different operating characteristics. It is extremely important in any application to follow the drive OEM's specifications. Some drives such as vector calculated speed control types do not allow any added capacitance or inductance on the load side of the drive and instead require removal any such devices prior to using an IGBT drive. I have yet to incur any expense associated with obtaining the services of a factory authorized, competent application engineer for preliminary evaluation to determine drive type suitability and performance specifications/requirements. Most (if not all) drive manufacturers are more than willing to invest initial investigatory engineering expenses in return for gaining a satisfied and well paying customer. My advice would be to decide which vendor type of drive best fits your application by inviting at least three different applications engineers from reputable drive manufacturers such as GE, Toshiba, Allen Bradley, ETC. to do a hands-on evaluation of your location. Ask for their summation in written form along with pricing, availability, cost, and warranty details specific to your identified needs. By the time you get through 3-4 walk-through evaluations with different vendor reps, you will be much better informed to make a good decision. It is very wise to be sure that you add at least 10% capacity to the recomended drive horse power to allow for unexpected operational dynamics in your system and to keep form operating a drive outside of it's design limits.

Good luck and remember:

The low price paid for any bargain is soon forgotten when faced with the cost of poor quality and performance.

I have had several vendors in and only one made comments about the cable run and went as far as to say. the run should be in conduit and the longer runs should have dv/dt filters. He indicated that the 20% in cable tray would need to be seperated from other cables and place dividers between them and the other cables. The application is a cooling tower fan with variable load. this application does not require a vector drive or any other specialty drive.

What is this 10% over rating of drive Hp. What happens when a drive runs at design limits. Can they handle a motor with a 1.15 sf and momentary overload conditions

thanks.

The desire by managment to find a way to do it concerns me, because the expense of drive rated cables and motors would stop the project. Drive rated cables and motors are the correct way to proceed.

Thanks

Most variable frequency drives available today are capable of providing a minimum of 160% of the rated motor starting current based upon HP demand without using a speed feedback circuit. In many cases by adding a tachometer type feedback to the controls, the output current of the drive can be increased up to 200%. However this magnitude of current is only available for initial starting requirements and only for a few seconds (40-90 secs).

Operating a drive at 100% of it's rated capacity for extended periods of time will cause premature failure.

Using existing non-drive rated motors and/or conductors yeilds aproximately a 10% increase in the demand placed upon the drive while decreasing the available horsepower and starting torque (rule of thumb).

Also the length of the drive-side supply conductors, their location/routing (in metal conduit or cable tray or suspended in free-air) and the type of shielding on the outside of the conductors determine whether your existing cable/conductors are suitible for using with VFD's and whether they have to be routed in separate cable trays and/or conduit.

All of these dynamics must be properly identified and factored into the equation before you can make an effective decision.

A good way to separate the competent and reliable vendors from those that are not, would be ask each vendor to supply a detailed application outline complete with a peformance warranty pertinent to your specific application in hardcopy form.

Be sure that the document contains/lists all of your system dynamics accurately, and specifically lists what existing or new equipment is required for proper operation to satisfy the warranty requirements. Otherwise the vendor will have a "loophole" for escaping liability.

Note: If they are not willing to committ in written form, most likely their quality of equipment performance and/or competency would be very suspect.

Basically inverter drives are suitable for speed ratio of 1:10 on the lower range from synchronous speed and higher speeds above synchronous are limited more by mechanical considerations.If a drive is fitted on an existing motor it is better derate the motor at least by 10%.If it is going to run constantly at lower speeds it is necessary to ensure temperatures do not go beyond specified limits like if your motor is having class F insulation and ambient is 40 deg C then temp under any operating condition should not exceed 145Deg C or temp rise over ambient of 40 is 105Deg.If ambient temp is low in your area that much advantage can be taken on temp .rise. As regds cable and its length it should be selected to ensure a that it can carry the reqired current and enure that rated voltage is available at the motor terminal.Inverter duty motors with external cooling arrangements are available which ensures sufficient cooling ven when motor continues to run at low speeds.

Krusher12,

You can balance two or three factors against each other. With higher carrier frequencies (and less audible motor "singing"), the cable length should be correspondingly shorter. With motors that are not rated as "inverter duty", their sensitivity to voltage stresses (from the drive's carrier frequency) will be greater, and will increase as the carrier frequency is increased. Also, higher carrier frequencies can reduce the maximum power rating for the IGBT's. Put all this together; it says to me: use the lowest carrier frequency you can.

When I was installing many drives in pumping stations, we always installed 3% reactors on the load side, but usually put them immediately after the drive (not at the motor). Nobody was recommending putting them at the motor, and usually there was more space or a better environment at the drive. Since the filter reduces the higher frequency harmonics more than the lower ones, it will also reduce the amount of electronic noise being radiated by the cables downstream from it. I think the drive location for the filter is better.

We paid a lot of attention to anticipated drive life, and usually specified drives to be rated for the motor service factor amps and also rated for no shortening of lifespan while operating in an environment with an elevated ambient temperature. Yes, this made them oversized, sometimes by as much as 40+%. But we wanted reliability. Sales reps hated us, but we found Baldor, Toshiba, and Schneider (SquareD-Altivar) to all be cooperative. Temperature is your killer, so giving yourself a margin at the drive can help a lot.

I assume that you will use an appropriate analog signal for drive speed control, and have some method for recovering automatically from drive or motor failure. If your application has multiple fans on the cooling tower, are you going to vary the speeds of all at the same time, or shut some off and vary others? A previous thread in CR4 had some discussion on this.

Regards--JMM

The drives come with 3% filters, but the manufacturer recommended 5% filters for the longer runs. I think this will help with higher frequency harmonics and provide cleaner power to the motor. This will help with life expectancy of the standard motor due to less stress on the insulation. I still expect a shorter life expectancy due to the use of the VFDs, but this will maximize it.

The VFD's are rated for 110% of max current for 60s. This is still below the 1.15 service factor motor current. Running the drives at rated currrent will ultimatly shorten the life expectancy.

Due to budget constraints we will not be able to use a plc for control. Looking to take analog signal directly from a temperature probe. May go with speed points instead to prevent hunting from the analog signal.

The fans would all be running at the same time and same speed, as determined by cooling tower heat load. This will maximize the use of the cooling towers capacity, plus provide tighter control of exit water temperature. I expect to save money on the reduced Hp requirement for most of the cooling loads

Thanks

Krusher12,

5% filters--good.

Sometimes you can use a single drive for multiple motors, if each one has its own running overload protection. This might allow you to have a higher reserve capacity in the drive and thus a longer operating life. Check the manufacturer's specifications.

Temperature sensor as analog input for speed control--good. This can directly control one drive with the necessary scaling functions being programmed into the drive's setup parameters. The remaining drives can then be wired as slaves from the first, using drive speed to set its analog output. . . .

Of course, during commissioning, you will need to watch for any speed ranges to avoid (due to harmonic vibrations in the cooling tower), and will have to set the minimum speed (so the motors can be cooled adequately). You probably will also find that the actual full-load current for the motor, when running at its nameplate frequency, is less than its nameplate current (because the drive is adjusting the motor voltage downwards some to keep the power factor as high as possible). This can lead to increased drive lifespan or could allow a slightly higher maximum frequency (to give greater cooling tower capacity).

If you try reading motor current with a clamp-on meter, remember that the waveform is non-sinusoidal, so a "true-RMS" type meter is mandatory. If you really want to explore the effectiveness of the load reactors, you can borrow or rent a frequency analyzer (Dranetz is one of the manufacturers of these). It will give you specifics for each harmonic in the waveform before or after the reactor at each monitored output speed (base frequency). Checking this at three or more different speeds is necessary because the harmonic frequency spectrum will be different for each base frequency and load. Monitoring at the motor terminals and comparing this to the frequency spectrum at the load terminals of the reactor will also help to establish the significance or insignificance of the cabling length or type.

A real "eye-opener" is to monitor the input to the drive and look for voltage transients, etc. I've seen spikes as high as +1500v superimposed on the 480v sine wave.

I like your plans and preparation. --JMM

Sorry about the late reply. Thank you for the excelent responses. The detail of your reply is very appreciated. this gives me a good starting point and additional food for thought. I was unaware of any vendors that provided the testing for harmonics. this is very interesting and would love to performe the tests. If I could convince management. First I have to get the project approved.

Thanks