Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations The Obturator on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

MV MCC at VFD Output 3

Status
Not open for further replies.

ThePunisher

Electrical
Nov 7, 2009
384
HI folks, happy new year!

We have couple of series booster pump motors in the past that are controlled by a single VFD. Since the pumps will be running at the same speed set-point, it was practical to connect them into one large VFD to save costs on space allocations (and building), auxiliary power supply demand and circuits, operation and maintenance costs (OPEX). The current design indicates that the motors are connected to a supplier designed MV distribution equipment, installed as part of the VFD assembly, with MV switches and motor protection relays (MPRs). The CTs are located in each switch outputs and is wired to these individual MPRs along with the motor RTD cables. These MPRs interfaces with the VFD controller. The only DRAWBACK we see is that when one motor is at fault or at over-temperature, the entire VFD shuts down thereby stopping all the series motors and hence disrupting the process operations.

TO ELIMINATE THIS DRAWBACK, We are intending to consider and explore the option to have a standard MV MCC at the VFD output instead using MV switch-contactor combination. The MV contactor will be vacuum interrupting type and will be controlled from a separate 60 Hz, 120 V circuit from a UPS instead of deriving the control power via individual MCC feeder CPTs. The MPRs will be powered from a separate 125 Vdc power source. In this idea, the MPR and contactor will "race" or trip instantaneously by the MPR and hence isolate the faulty circuit or motor before the VFD decides to shutdown the entire process train. Our motors will be the 16 pole types (without gear boxes) and hence, we can operate the continuous process speeds within 30~60 Hz.

My concern would be the following:
a) Is there an issue operating the 60 Hz CTs on non-rated frequencies and if there is a continuous frequency range where they can operate properly?
b) Can I operate the contactor 120 V circuit from a CPT connected to a supply that is in variable frequency operation?
c) Is our design option sound hoping that others here did this in the past and if so, what are the lessons learned?

We will be contacting the VFD and MCC suppliers for guidance later but I am hoping this option is not that new and folks here already did it. I ATTTACHED AN OVER-SIMPLIFIED SKETCH TO SUPPLEMENT MY QUERY.

Regards and thank you in advance.
 
 http://files.engineering.com/getfile.aspx?folder=2f7cc3ec-89fe-485b-95a7-4c914ac85d70&file=VFD_CONFIG.png
Replies continue below

Recommended for you

I have a suggestion. It seems like you're going to a lot of quasi-exploratory effort just to solve the one-motor shutdown issue.

Other possible solutions:
1) What about having the already required individual motor protection relays remove the overheating motor without having the VFD shutdown. You might want to have a simple voting scheme where, say, more than two motors shutdown causes the VFD to shutdown.

2) Change nothing except to prevent the individual motors from ever tripping on overheat as that is unreasonable abuse of them anyway. This is an avoidance move that would actually pay you back better than any method so far mentioned. It will increase the lives of all hardware. Prevent overheating by either increasing cooling or actually measuring the motor temps and alter the process proactively to prevent individual shut downs from ever needing to happen. A cheap PLC could do this task for a couple of hundred dollars.

Keith Cress
kcress -
 
HI Keith, thank you for your prompt response.

On your suggestion #1
Is this suggestion assuming that each of the motors are connected to a MV MCC with switch-contactor arrangement? In my attached diagram, my intention is to have the MPRs directly trip the contactor when there is over-temperature and over-current. However, it will still send a signal to the VFD to shutdown but with a pre-determined time delay. If this is an over-temperature, the MPR may be programmed to have a High and High-High with time delays wherein High will be alarm only to allow the operator to make adjustments. High-High will trip to the contactor instantaneously and will send a signal to the VFD to shutdown with time delay to allow the contactor clearing first (back-up on a bigger protection zone). The voting scheme is a good idea but has to be collaborated with process and operations.

On your suggestion #2
Is this suggestion assuming that each of the motors are connected to a MV MCC with switch-contactor arrangement?
"Change nothing except to prevent the individual motors from ever tripping on overheat as that is unreasonable abuse of them anyway"...if the motors are already over-heating, that is already an unreasonable abuse of them, isn't it?

If I don't change anything, a fault on one branch circuit or motor by overcurrent will shutdown the entire VFD and consequently the process. We already have that to begin with and are trying to improve the process.

Making necessary loading adjustments to reduce any expected over-temperature would require some multiple stages of alarms to warn the operators or develop an algorithm to inherently adjust by PLC is a good suggestion. But perhaps, I would still maintain the High-High trip to ensure that the specific motor trips out if the operator runs out of time.

For either suggestion #1 or 2, if the overcurrent protection is used, the MPRs will be individually set to have an overload protection of 100% FLA and a short circuit of 150% of total VFD FLA as the VFD will tend to current-limit on a short circuit to 150% for time-delay.
 
Hi TP;

I was really only speaking to OTemp not SCircuit failures which are certainly a possibility and need to be considered too as you've done.

Yes I'm assuming each motor is "MV MCC with switch-contactor arrangement" if you want to have any single motor drop out.

Your High/High-High scheme is a lot better than what I was imagining of just dumping off a motor every time it overheated. Still humans are so much more fallible than a PLC that's watching things, as noted by your even having to have the High/High-High scheme in the first place.

If you expect operators to make operational decisions about what actions to take based on pending overheats or even short circuits then you should be able to distill those same operational actions into automatic actions done by a non-organic control system. Presumably someone would have to write an operations manual which is most of the control system behavior distilled.

But why stop there and not proactively prevent the OHeat problems all together. I would be setting up the system to react automatically to a pending overload by a shift in the operating point. Once that was functional I'd use the monitoring to feed back to whatever it is that's causing the overload in the first place to prevent it, as it results in reduced throughput of the entire system while causing the over-heats. It could be to slow a feed conveyor or even a red light telling a human loader to stop or reduce the loading rate because the ore is heavier or whatever. It would be easy to provide a human loader with a light scale showing three, four, or five levels of loading.

"Change nothing except to prevent the individual motors from ever tripping on overheat as that is unreasonable abuse of them anyway"...if the motors are already over-heating, that is already an unreasonable abuse of them, isn't it?

Dang straight! It's very hard on large motors because they are so thermally massive. Throwing them into stop can mechanically shock the entire system, leaves them stewing in all their over-heating with possibly even hotter internal points developing due to loss of all airflow. I'm not clear on how losing one motor doesn't immediately push all the rest of the motors closer to their own OT limits. Seems like a classic setup for a cascading nightmare. That's why I'd avoid having it ever happen by design of a relatively simple control system. You have the expensive part already - the VFD.

Keith Cress
kcress -
 
I look at this a little differently.
First, why is a motor overheating. Is it being overloaded or is it begining to fail.
Let's look at overloading, which is probably the most common issue.
With more than one motor on the VFD, one motor may be overloaded without the drive being aware of the issue.
You need some way to monitor the load on each motor and reduce the load on any overloaded motor. You may be able to throttle a discharge valve to reduce the load.
I would first try to prevent overloading rather than letting a motor overheat and then trip it off.
Which would you prefer; Reducing your through put a few percent to avoid an overload, or letting a motor overheat and trip off and lose about 30% of your through put?
I can think of a couple of possible ways to do this but the person who solvs the problem may have to be familiar with your entire process and equipment.


Bill
--------------------
"Why not the best?"
Jimmy Carter
 
How well will the VFD handle part of the load suddenly switching off? Some VFD's would not like you doing that very much. You also can't re-start the tripped pump without stopping the VFD, so the process has to shut down to bring it back online.

You can't use a CPT on the output of the VFD. The output voltage and frequency of the CPT will track the VFD output. CT's with enough iron in them (ie high class rated ones) should be fine over the 30-60Hz range.

My advice would be that you should put a VFD on each motor and load than can operate independently. A 12,000hp VFD will be very large and expensive so the cost savings of a 12,000hp VFD and the switchgear might not be that much compared to 3 x 4000hp VFD's. Trying to save money by making control schemes where you pick the loads and then run them with a single VFD creates a complete pain in the ass on the operation side. Then, you also have a single failure point and when the VFD has an issue nothing works. Do notice that I posted "when", not "if" because that single VFD will fail.
 
First of all I thank you for your valuable posts (stars to come later)...

The process operations of these pumps are currently being defined and I would ensure that I will be reviewing them later once our process team (in collaboration with operations) issues a draft for internal review. However, based on the initial understanding our process team provided to me, here how the process is initially perceived.

The 3 x 4000 HP pumps are connected in series in a train at 3 x 100% capacities. They pumping coarse tailings slurries that is being discharged into an open pit creating beaches and pond that is dewatered for the purpose of creating sand dump. These pumps are designed to operate at similar flow rates and pressures for the purpose of attaining a specific acceptable narrow range of output. These coarse tails are obtained from a main plant that also governs required flow rate and pressure based on its process system. Hence, these pumps relies on what's being thrown at them as part of an overall process system. In this regard, if the pumping per train losses capacity, it would ripple all the way up to its source and operators or the overall process logic may have to re-adjust with the aim to maintain the required output at the end. There are three trains where each would have 3 x 100% pumping capacities and there will be cases where 2 out of three will operate up to a worst case of all three operating depending on the upstream process output condition where these tailings wastes relies on.

Hence, by process design, they are not allowed to "hold back" or reduce pumping capacity because one motor or one train suddenly reduced its output pumping capacity (either operating at 2 x 100% or 1 x 100%).

Our initial design intent was to have individual VFDs for each 4,000 HP motors or 3 x VFDs per train. We obtained and compared the required VFD physical foot-prints, auxiliary power requirements and equipment costs (including heat exchangers) for single individual VFDs versus one combined VFD and the significant cost savings are realized for the latter.

Since the three series pumps would operate at 3 x 100% and are intended to operate collectively to attain required output capacity REQUIRED collectively and not as to have standby or spare systems. Having one VFD per train is being explored as an option.

Depending on the eventual process operation criteria, I would think that using contactors to trip one fault motor on a over-temperature High-High or one motor branch circuit short circuit would both "buy time" for the operator (or process system to do necessary adjustments rather than one whole train shutting down).

On an over-temperature, a high alarm say would go ON and gives the process operator "some time" to make necessary adjustments (that we will find out later) process-wise (say start up train 3, etc.) and trip out the faulty motor when the alarm reaches high-high level. There are two sets RTDs per motor winding (total = 12 winding RTDs) and 3 bearing RTDs. One set of winding RTDs (3 x RTDs) will be wired directly to the MPR and analog temperature signals can be interfaced to the DCS via MPR OR we can wire the other winding RTD set independently to DCS and the operator can be provided with both over-temperature readings from RTD and MPR with a programmed selector switch in the control room.

I hope I did not bore anyone with the lengthy explanations above.
 
And by the way, the 120 VAC contactor control circuit will be separately supplied from a fixed 60 Hz supply.

The motors are expected to be either the 12 pole or 16 pole design (our client hates gear boxes) and hence the pumping power requirement may be within the 30~60 Hz speed range.

It was mentioned that when a VFD is operating 3 x 4000 HP and suddenly a load rejection happens due to a high-high over-temperature or a short occurs on one motor terminals causing it to suddenly operate from 3 x 4000 HP to 2 x 4000 HP; would create issues to the VFD. If the VFD is sized to 3 x 4000 HP, is sudden load rejection going to be detrimental to the VFD electronic circuitry?

I would like to understand better and your inputs are highly appreciated.

I will start discussions to reputable VFD manufacturers like Siemens and Nidec later..but a good heads-up would be wonderful.
 
Motors that are well built, properly applied and not overloaded seldom overheat.
In applications where operators are able to overload motors, motor overheating may be frequent.
You may be attempting to treat the symtoms rather than the cause.
With three pumps properly connected in cascade, the load sharing between the motors should be quite good.
There should be no reason for only one of the group to overheat due to overloading.
If the mechanical group has planned three pumps in cascade there may be a good reason.
I suspect that the pump curves are such that with one pump out of service, you will slide so far down the pump curve that you will get far less than the 67% that simple arithmetic suggests. You may have exceeded the available head and get zero through-put with only two pumps online.
I had an issue when the pump head was exceeded and so little fluid was delivered that churning in the pump actually boiled the sewage in the pump. Fortunately the situation was corrected before the seals were damaged. Some paint was burned off the pump due to the heat developed.
Running a pump past the end of the pump curve is similar to running into a closed valve as far as the pump is concerned.
Electrically, tripping out one motor may be a very good idea.
BUT
When you step back and take an overall look at the whole installation, it may not be worth the effort to trip only one motor.
This is not to say, don't monitor the motor temperatures. For an installation of that size, I would consider monitoring the current as Itsmoked suggests and also monitoring the temperature of the motor, the pump and the bearings.
Continuous trending the bearing vibration may be a good idea also, and may be combined with temperature monitoring.

Bill
--------------------
"Why not the best?"
Jimmy Carter
 
Was that explanation written by a lawer?

What do you mean by series? Typically, when pumps are connected in series all the pumps are required to produce the required head pressure. Meaning you don't just shut down one pump and expect the others to still work through the shut-down pump. With series pumps it is pointless to have breakers, contactors or even switches between the VFD and the motors because they all need to work. Maybe fuses if you wanted to "meet code", but the VFD will trip-off due to a short before medium voltage fuses on each motor will blow.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor