What order of magnitude of power draw difference are you talking about? In my experience with double drive mills, there is always a master and slave set up, so one motor will invariably draw slightly more current than the other (I am referring to a WRIM type double drive) this then shows up as a power difference. Typically a power difference of -5% is normal. Charge motion should not greatly affect either drive, as once started, this motion is fairly stable. Feed surges or other process issues should also not greatly affect the current draw of each motor separately, rather if you get a surge, both motors should ideally increase their current draw simultaneously.

There are two GE motors for SAG mill but unfortunately sometimes current draw exceeds more than 300 A in one of them and makes SAG mill shut down. I was not sure about process issues. Thanks a lot :-).

Interested in getting update from you:

What is the power draw difference between the two motors?
Is the difference consistent or fluctuate?
Is the mill normally operating at the motor power draw close to the shutdown set point? How close is it?
When the current draw exceeds the set point for one motor, what is about the other motor?
Is the power voltage consistent?
Was anything process wise happening when the current draw exceeds the set point? Such as any change of water addition, feed load, recirculation (through pebble crusher) load?

Answers to the above questions can probably give clues to the cause of the problem. Please update us if you've resolved the issue, what you did.

Average of difference between Current draw for both drives is 20 (A) in case of forward rotation with standard deviation about 22 (A). As you can understand from high standard deviation, the difference is not consistent but in 93 % of operation times, drive number 2 draw more current in both forward and reverse rotation. Power draw is about 7300 (kW) (max of Power draw is 8200 kW). The voltage always is consistent and there is no significant process change during SAG mill operation.

Are there clutches on these drives? If there is it could be a control or mechanical issue. Some manufacturers install load balancing systems for this very reason in the case of drives with clutches the system will let one of the drives slip by pulsating the clutch air until the two drives are loaded equally or within a very small deferential.

When you say the difference is not always consistent, can you tell when, how much and what pattern is the deviation? When the high power motor breaks the setting, does the current rise gradually or suddenly? Also like to get your answer on the above Question 4.

Look at the brushes on each ring to see if they have the same wear and the plastic deposit from the brushes show uniform color and consistent look.

Have you cleaned the rings and installed new brushes lately? The brush milon (sic) insulation deposition sometimes is not uniform which can lead to differential current capacity between brushes and between drives. You must clean all rings and replace all brushes simultaneously or the problem will get worse.

If your climate is dry, you also might find a small pan of water placed on the bottom of the brush box helps. We found this so in the Arizona and Chilean desert.

To minimise the differential between WRIM motors you can try carrying out an electrical phase alignment of the rotors. We carry out the procedure each time a motor is changed. The phasing is conducted by connecting a 3 phase supply to the motor stators (nominally 415V) with the rotor connections open circuit to facilitate voltage measurement. With one of the motors decoupled from the gearbox the rotor is turned and the voltage measurement:

Connect one lead from a hand held voltmeter to K phase of the slip ring busbar.
Connect the other lead from the voltmeter to K phase of the slip ring busbar on the other motor.
Ensure connections and area is safe then turns 415v power on, proof 415V is available.
Rotate the shaft on motor B to align each bolt hole (1 to 8) with reference bolt hole on gearbox coupling.
Measure the voltage change down to the minimum voltage. Record the measurement in the table below and repeat for each phase (K, L, and M). This ensures phases on both motors are connected to the slip rings in the same order.

Note: The bolt hole with the voltage difference between same phases is lowest is the position the motors are to be coupled to the gearbox and locked in. This is the magnetic center for the electrical alignment.

Post a fitter at drive end of motor “B”.
Mark on the motor hub and gearbox hub with the position they need to be coupled with a white paint marker. This is to ensure correct position in case the gearbox is turned before final coupling.

"Average of difference between Current draw for both drives is 20 (A) in case of forward rotation with standard deviation about 22 (A). As you can understand from high standard deviation, the difference is not consistent but in 93 % of operation times, drive number 2 draw more current in both forward and reverse rotation. Power draw is about 7300 (kW) (max of Power draw is 8200 kW). The voltage always is consistent and there is no significant process change during SAG mill operation."

I understand from this that you have an approximate 20A current draw difference between the two motors? I also assume that 7300kW is the TOTAL mill power draw, or is it the power draw of EACH motor?

I believe that you have the Master-Slave arrangement set up. You state you have 8200kW installed total power. That shouldbe 4100kW per motor. A WRIM of that size would draw (an estimated) FLC of 380-420 Amps. So a 20Amp delta between them is approximately 4-5% of the total current draw and typical of the Master-Slave set up.

I like above suggestions to get the motor phases balanced and check that both motor slip rings are connected in the same order.

I also agree that slip ring and brush maintenance is vital, I am aware of some large (6MW) WRIM's that had major issues (i.e.; FLC flash over) that was solved with brush compound changes and thorough slip ring cleaning / maintenance.

Another area that may be worth checking is the control of the switch gear on these motors. You said earlier that the mill will trip when motor current exceeds 300A. This is unusual if the motors are 4.1MW units as they should draw this level of current anyway.

It could be that the trip point for the mill is set too low? You may also want to check the manufacturer’s no-load current draw of the motors.

Thank all of you guys for your very good advices.

7300 kW is the total power draw (both dives together). Increasing current draw is not suddenly but gradually. When current draw reaches 300 (A) in one drive, the other one draws 283 (A).

When you find out what this issue is please let us know I'm curious to know what actually turns out to be the cause.

In the typical twin WRIM configuration, there are two WRIM, 2- gearboxes, and 2 pinions driving a common ring gear connected to the mill shell. Run out occurs at the interfaces between the pinions and the ring gear, Since these are distanced from each other, the pinions cannot share load equally so there is a load swing between them during each revolution of the mill, and this accounts for a typical 5% load swing between them.

This is typically compounded by the run out in the individual motor gearboxes. Which again generate load swings between motors and gearboxes.

Because twin WRIM motors are undamped in their relationship with the gears, there is a simple harmonic motion set up in the drive train which produces uneven wear in the gear trains and damage to the gears, over time. So damage to the drive system is inevitable.

By contrast, the GE patented Quadramatic twin drive system will deliver precise load sharing between two low speeds Synchronous motors, driving through a common ring gear.

I agree that if there is a difference between centres of one pinion-gear interface to the other, then there is unequal load sharing, and this may cause the motor power to increase on one drive as opposed. However, to me this unequal load sharing would be apparent from start-up and would probably cause a mill trip on overcurrent on one motor during the first few seconds of a mill start. He mentions a gradual increase in motor current until one reaches 300A and trips the mill. Checking the mechanical set up (centres) and clearances (backlash) would be advisable. As to "run out" or backlash in each gearbox, this would be taken up at start up when the pinion is loaded by the gear. It would be very interesting to inspect the gear and pinions and compare the contact patterns. The key to me is the power supply to the motors must be balanced, and the switchgear should be common to both.

It does not hurt to look into the electrical system and try balancing the load between the two motors. I think the cause is more likely at the process side. Suggest comparing the power load curve with other measurements, such as load cell data, trunnion bearing pressure, acoustic whatever installed. You may find similar changes that can be indicators of process change.

Compare the power load change with the liner change history. If you find any trend on that, liner wearing can be a cause.

Ball load may be a cause too. Ball load can change because of feed rate change or wearing rate change. Ore property can also change hardness, clay content, etc.

There are many paper talking about SAG mill optimization that can give you clue how the process data changes the power draw.

Another difficulty with WRIM is the permanent resistance, including any tuning resistance plus the rotor resistance itself. Maybe consider strain-gaging the two drive shaft to measure the true torque and not any integrating (volts, current, and power factor) gages. Based on the torque, validating the motor performance at the given RPM, motor current and voltage, you can identify whether the permanent resistance step is balanced.

Good point - it may well be that one of the motors has insulation problems on windings in either rotor or stator, or other issues that are influencing its performance. There may also be coupling or alignment issues.

I also think you may be onto something - a gradual increase in current draw on one motor may well be linked with an increasing mill load (balls / clay / ore / water etc) or surges in feed that will end up overloading the mill over time.

Unfortunately I’m going to leave this concentrator for a while and taking apart in another project somewhere else but I will be in touch with this issue and follow it. I will make you aware about conclusions.

Maybe an FFT might identify if there is a correlation with rotation of pinion or ring gear.

This has been a very good read. A lot of input from various sides - mechanical, electrical, process!

In my opinion, current imbalance on motors of a twin Pinion Mill Drive train is caused by inaccurate load sharing. Power supply to the motor has no effect as it is passive and power is drawn as needed.

Since this is a WRIM and LRS set up (no VSD with active load sharing or Quadramatic) the motors rely on natural load sharing. For this the resistance of the rotor circuit must be somewhat identical between the two motors. This resistance circuit comprises of a number of “sub-resistances” in series – Rotor resistance + LRS resistance + Brush resistance and resistance of the contact surface between Brush and Slipring. You advised that the problem occurs during operation, not during start up so AFTER the LRS has been bypassed so we can disregard any uneven resistance in the LRS (which can still exist but it would only cause unbalanced current during start) .

I therefore tend to agree with him and would start investigating the brushes and the surface area to the sliprings. In order to determine whether the rotor resistances are identical, I further agree that one should start carrying out independent torque measurements on the motor shaft (using strain gages or similar) to see how balanced the motors are. Rotor resistances are never identical (even on motors with identical KW’s) but this can be mitigated with trimming resistors.

I like comment in regards to Process. I believe that changes in process plays a role. Now, we all know what a typical torque/speed curve of a WRIM or SCIM looks like. In the area of the operating point (about 99 % of the synchronous speed) the operating point moves along a steep curve. If the motor is overloaded, the speed is inevitably reduced, causing the operating point to move UP the steep curve, drawing more current and developing more magnetic field and torque. That extra torque helps increase the speed which causes the operating point to move down the curve again– that backward and forwards motion or “balancing” happens until a stable operating point is found.

Every time the load in the mill changes, the balancing has to start again. The steeper the curve, the more current is drawn for even the slightest reduction in speed. The steepness of the curve is determined by the slip of the motor which in turn is determined by the rotor resistance. Adding more resistance (trimming resistors - as suggested) to the rotor will flatten the curve and make current imbalances during load changes less severe. It will also reduce continuous current imbalance during operation and minimize mechanical wear and tear on pinions and girth gear.

As I understand it, this "balancing" would occur in each motor. Could it be the case that in the event of a twin drive that the motors are fighting each other, especially if the rotor circuit resistances are not equal?

It is correct that if you have two motors whose rotor resistance is not identical (which is often the case) and who are mechanically coupled either through a pulley on conveyors or a girth gear on mills, they will go on forever trying to find a balance (or fighting each other as you put it). However the resulting “load swings” are relatively minor (between 3-5 %) and are accepted by the Industry.

If you have a WRIM, you can try to add trimming resistors to balance out the rotor resistances and minimize load swings but if you have a SCIM, there is no way you can access the rotor circuit. In that case you need "smart" load sharing and for this you need a VSD. The VSD can be set up in Droop Control which basically compensates for that difference in resistance.

These days we often use a combination of Master – Slave and Droop control to do load-sharing on conveyors. On Twin Pinion Mill Drives above 8 MW however, he believes that Master-Slave or Droop control is not accurate enough which is why we apply an independent, superimposed Motion Controller for a dynamic speed error of <1 % of nominal speed.

Most of our large mill drives >5MW per motor are WRIM, however we are now seeing a swing toward SCIM and VSD's as the cost of CAPEX for VSD's is decreasing, and operations are most sensitive to energy savings. A set of WRIM's running at 80% speed being controlled by an LRS is not the most energy efficient way to operate. Our last two 'large' drive SAG mills are both using WRIM's connected to SER drives. This solution seems to sit in between WRIM+LRS and SCIM+VSD as a cost effective drive solution.

One word of caution with SCIM and VSD on mills, if you have a gearbox or gearboxes consider the failures seen in the wind industry, variable speed can wreak havoc on roller bearings (smearing effect)! To help increase bearing life they apply black oxide coatings. I say this because we have experienced this issue.

I haven’t heard from smearing before and looked it up on the SKF site.
“When two inadequately lubricated surfaces slide against each other under load, material is transferred from one surface to the other. This is known as smearing and the surfaces concerned become scored, with a "torn" appearance”.

VSDs can induce unwanted motor shaft voltages that lead to currents that circulate between the rotor and the motor housing via the bearings. These currents can harm bearings however, this has been known for a long time now and every 3-phase AC induction motor used for “inverter-duty” is fitted with a shaft grounding system which effectively is an earthling brush on the motor shaft that offers an alternative path for the currents to earth. In addition to that, larger motors are typically fitted with insulated bearings which eliminate this issue.

You mentioned gearboxesand there is the case that if the motor bearings are insulated and there is no earthing brush, these shaft currents will look for an alternative path and continue flowing via the shaft into the gearbox or even further into the machine where the flow to earth via the bearings. However, there are ways to prevent this as well.

I don’t know whether this is what you call smearing but if it’s something else, can you please elaborate on how the smearing on these bearings and VSD operation are related? We see a trend away from the LRS starter because of reliability issues and high maintenance requirements. It’s good to see that confirmed from a Mill OEM. In order to increase mill grinding efficiency and minimize maintenance, most clients decide to upgrade their existing fixed speed system to a variable speed system - whether this is with a SER drive or a VSD. The Idea is to keep the WRIM and simply replace the starter, however, since its mostly the LRS that is causing grief – I don’t see how a SER System alleviates the problem as it still requires the LRS.

Variable speed can cause rolling elements to slide instead of roll as designed when this happen there is a chance that smearing will occur just ask your friends in your drive department in Voerde (previously Felender).

I also think that as the costs for VSD systems decrease, plus that the technology is mature and becoming widely used on large machines, that this is why we are seeing more of them. However, even though LRS units are higher in maintenance, they are simpler to maintain and do not require a high level of expertise or external support, which serves well on some remote mine sites or at emergency breakdown situations. LRS's are simply easier to fix and maintain. With operators wanting to optimize recovery and maximize liner life, especially on larger mills, I can understand that going to VSD's is an attractive option. However, I think it does not make sense to fit a VSD system to a machine that will spend most of its operating time at a fixed speed, i.e. large ball mills. S.A.G is a different story and should have VSD as standard.

Even if a client decides to "upgrade" from WRIM+LRS to a VSD, they still have to replace the motor (as a WRIM cannot run under VVVF control), so it’s not just replacing the starter, unless I'm missing something?

Good comment and an important point. I also looked up "Smearing" and found that this can occur if the roller has a different speed in the unloaded zone than in the loaded zone and is caused by high acceleration of the roller relative to the ring, this causes sliding as opposed to rolling contact and results in smearing. I think the damage you refer to is caused by EDM - "Electrical Discharge Machining" as a result of stray or eddy currents passing through the bearings if not properly grounded. You refer to the wind industry, where the drive is in effect the opposite way around compared to a mill and the blades are driving the gearbox that in turn rotates the generator rotor to make electricity. From that point I can understand that the speed of rotation is not always constant because of the wind, this could lead to irregular acceleration within the gearbox / motor.

Whether smearing can occur in SCIM and gearboxes in mill applications due to Variable Speed operation remains unclear - as typically VSD's have a very precise speed control function which would eliminate irregular acceleration / deceleration events, and the modern SCIM's and VSD electronics have insulated antifriction bearings at both ends, good grounding and cable sizing. The use of proper conduction grease and proper grounding methods also helps to reduce stray eddy currents. Modern power electronics and the correct filtering equipment used in VSD's also help to reduce these system currents.

It would appear (from what I've read) that the main causes of smearing are incorrect mounting, insufficient lubrication and heavy axial loading. Also, to my knowledge, motors up to about 5-6MW is the limit for antifriction bearings, beyond that, sleeve bearings are the norm.

However I do agree with you that proper bearing selection in motors and gearboxes, plus good grounding / insulation methods should be employed for these large drives.