Soft Starter for 11kV submersible pump 7

[Loghman_ra]

Hello Fellows

I have a case to discuss,
we have a 11kV submersible motor , 1000kW.
it is asked to use soft starter for starting.
there is 2 limitation:
1- the submersible pump shall reach half nominal speed in less than 5 second which is due to mechanical requirement of the pump,
2- we are trying to restrict the starting current less than 350% or if possible 300%,

the motor/pump data is not finalized yet, some supplier propose VFD instead of Soft starter but we want to realize if SS is possible also or not?

then please share your experience for similar case.
BR

 

[Gr8blu]

A "soft starter" comes in many different forms, including the use of power electronics (i.e., a drive). A VFD rated for "soft start" duty will not operate the motor continuously; it is sized to run in the drive's overload region for the duration of the acceleration period, followed by switching the motor to the line (and thus providing a "dwell" period for the drive components). Depending on the pump/load inertia and the motor developed torque, an across-the-line start of the motor might take anywhere from 3 to 30 seconds, while pulling 5x to 7x rated current. If the max current limit is going to be held to 3x to 3.5x range, the time to accelerate will increase correspondingly (not quite a linear relationship, but enough for estimation purposes). Note that a "low inrush" pump motor is feasible, but may be physically larger (longer?) than what the normal rating might look like.

Converting energy to motion for more than half a century

 

[edison123]

If you have the money, then VFD with filters and VFD cables. No torque loss during low voltage start. No large inrush currents. Variable speed for variable flow, if you ever need it.

Muthu

http://www.edison.co.in

 

[waross]

2- we are trying to restrict the starting current less than 350% or if possible 300%,

Are you trying to restrict the motor current or restrict the line current?
There are reasons for restricting line current (Soft grid).
Motor starting current is highly reactive.
The reactive current may be supplied by capacitors with an expected reduction in line current from 600% to about 300%.
The capacitors should be cut out in stages as the reactive current decreases with motor acceleration.
However, with a soft grid, after consultation with the supply authority, consideration may be given to leave extra capacity on-line to hold the voltage up.



--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[jraef]

A VFD can definitely do it, because a VFD can provide 100% torque from the very beginning without exceeding the motor FLA, or even up to 150% for a few seconds if necessary (although not likely). With a VFD, the line current and motor current are the same all the way through the acceleration and the reactive current of the motor is controlled and provided by the VFD, so the current is almost all torque producing current, which facilitates the fastest possible acceleration if line current is a concern.

As mentioned, any “Soft Start” means applied to the motor will result in an increase in acceleration time. A Soft Start trades the peak current for acceleration time. The relationship is not linear so predicting it is difficult at best, without using sophisticated software that can perform what’s called a “Transient Motor Starting” (TMS) analysis. If you have access to software such as ETAP, SKM Power Tools or something like that, they can predict the possibility of accelerating within 5 seconds at a particular current limit, or conversely tell you what the current limit setting must be to accelerate within 5 seconds. These software packages are expensive, but nowhere near the cost difference between a Soft Starter and a VFD in 15kV class equipment. You will also need a COMPLETE motor and pump data set, no guesses if you want accuracy. Even then, the risk is that you buy the software and discover that a Soft Start will not work, then have to buy the VFD anyway… An alternative is to find a Consulting Engineer with this software and experience in using it, and pay them for the study.

As a side note: is the motor already purchased? If not, for 1000kW I would suggest using a lower voltage such as 6600V (7.2kV class), because the cost of the VFD or Soft Start will be significantly lower. Once you get above that, most VFDs will be this lower voltage anyway, with transformers on both sides to step down, then step back up if the motor is 11kV. The literature on them might not show that unless you can see the one-line diagram, but power electronics above 7.2kV class are rare and extremely expensive. With Soft Starters there are 15kV class available, but the cost of the switchgear associated with it is significantly higher. Of course if the down hole distance is extreme you may have no choice but to do it at 11kV, but it’s worth looking into.




" We are all here on earth to help others; what on earth the others are here for I don't know." -- W. H. Auden

 

[EdStainless]

How deep set are these? The real question is how long are the leads?
The inductance in long cable runs is not to be overlooked.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[Loghman_ra]

Thanks Gr8blu,
maybe it is better from now I use RVSS instead of SS. so we are trying to not using VFD.
of course there will be bypass switch after start-up.

Thanks edison123
as mentioned client asks for RVSS.

Thanks waross (
yes your guess is correct, line current needs to be limited due to our grid.
also I saw in some RVSS supplier product, there is capacitors by default ,
but I am not sure if these capacitors can help during transient time (start-up) or not? since also client ask to keep PF at 0.95.

Thanks jraef
yes , for sure we will do calculation with ETAP,
motor not purchased, we are at the stage to see which solution can be possible , some VFD supplier says (no evidence) using RVSS is not possible to fulfill the requirement( 5 sec start up + 300 % starting current), but we should should justify the client if RVSS is not possible anyway.
Good tip to use 6 kV motor , even the VFD supplier normally offer based on 6 kV motor.

Thanks EdStainless
distance between main 11kV switchgear and motor location is around 60m ,

 

[LittleInch]

Also try to understand what exactly the motor load at startup is and challenge the data because at the moment it doesn't make a lot of sense to me and implies full speed in <10 seconds which is still pretty fast if you're trying to keep below 300% of FLA. Also getting what the actual torque speed curve is for the driven element (the pump)

In these sorts of situations you often get people giving you unrealistic requirements or "standard" data and in reality and with a change of start up can reduce the loads considerably.

So challenge everything that comes from others as this is a sizeable submersible motor and pump and your limitations are quite significant to try and meet.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[edison123]

Weak grid + Restrictions on starting time and starting current + Poorer starting torque with RVSS + 1000 KW motor 60 Meters down a hole

= 6.6 KV VFD (which is also a RVSS) + Smooth start from zero to full speed with no current spikes + Variable pump output with possible power savings + Fewer operation issues.

No brainer.


Muthu

http://www.edison.co.in

 

[Loghman_ra]

edison123 (Electrical):

Yes , VFD is a solution but costly and also more consideration rather than a RVSS (no frequency change)!
we are trying to find a solution based on RVSS ( no freq. change)

https://vfds.com/blog/what-is-a-soft-starter/

 

[Valvecrazy]

Usually because of the Kingsbury type thrust bearing, a submersible needs to get to at least 50% speed very quickly. While a SS will reduce the amplitude of the starting current, starting a centrifugal pump against a closed or almost closed valve can greatly reduce the duration of the start. Then the valve can be opened as fast or slow as needed.

 

[Loghman_ra]

Thanks Valvecrazy
good Tip!

 

[Marke]

Both the VFD and the soft starter can be used up to 11KV.
If you are using a soft starter, it pays to look at the high slip performance of the motor, especially the Locked Rotor Current and the Locked Rotor Torque.
As you reduce the start current, you reduce the start torque by the current reduction squared, so a modern high efficiency induction motor with a very high locked Rotor Current, and a low Locked Rotor Torque will require a higher start current than an older motor with a higher Locked Rotor Torque and a lower Locked Rotor Current. - The motor selection is very important for good results.
With a VFD, the motor always operates under low slip conditions, so the high slip characteristics are not relevent.
The advantage of a VFD are :
[ul]
[li]always runs under low slip conditions so the start current is low.[/li]
[li]can run at speeds other than line speed. - slow down when full speed is not needed. [/li]
[/ul]

The disadvantages of a VFD are :
[ul]
[li]Higher power loss - typically 2.5 - 3% losses in the VFD always while running[/li]
[li]High input harmonics, or higher losses if harmonic mitigation is fitted. - perhaps as high as 6% if harmonically mitigated. [/li]
[li]EMC and EDM problems due to output waveform. [/li]
[li]Higher capital costs and shorter life expectancy compared with Soft Starter. [/li]
[li]Potential insulation damage over time[/li]
[/ul]

The advantages of Soft starter are :
[ul]
[li]Much higher running efficiency and can be bypassed to virtually eliminate all losses. - lower running cost at full speed. [/li]
[li]Lower Capital Costs [/li]
[li]Typically, longer life expectancy [/li]
[/ul]

The disadvantages of Soft starter are :
[ul]
[li]Higher starting current[/li]
[li]Dependent on high slip motor characteristics.[/li]
[/ul]

There are many tales about the energy that is saved on all motors with a VFD. The motor efficiency does not improve with a VFD compared to a soft starter, but the energy wasted in the driven load can sometimes be reduced by slowing the load down. In many cases, because the work done is reduced, the motor runs for longer.

There is a genuine case for each technology and it depends what the application problems are as to what is the best solution.

For steep curve pumps with variable flows, there is energy to be saved by slowing the pump for low demand.
For flat curve pumps. slowing the pump reduces the discharge pressure, so it will always run at a fixed speed with varying flow. No energy to be saved, just 6% additional energy to be wasted.


Mark Empson
Advanced Motor Control Ltd

 

[waross]

You may have misunderstood my suggestion.
I was not talking about running power factor correction.
A a large motor has a starting power factor in the order of 20% near locked rotor.
That is almost all reactive current.
Some of this reactive current may be supplied by capacitors rather than drawing it from the grid.
I would consult with the motor manufacturer's engineers, (not the sales engineers) as to the reactive current that the motor may be expected to draw during starting and opinions as to the feasibility of using capacitors to supply reactive power during starting.
The starting capacitors should be cut out as the motor accelerates.
Basically, the motor is started DOL, but not all of the starting current is drawn from the grid.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[Marke]

Totally agree about the starting reactive current, but the power factor changes with the slip of the motor and if you are able to correctly compensate a rapidly moving target, you can significantly reduce the intial start current. You need a very high speed compensation loop though, as a pump accelerates pretty fast!
The PF generally follows the slip in a linear fashion, max KVAr at zero speed and min KVAr at max torque point on curve, around twice the full load rated slip.

Mark Empson
Advanced Motor Control Ltd

 

[waross]

Is vacuum contactors and point on wave switching to avoid energization transients the way to go, Marke?

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[Marke]

Hi Bill

I would be looking at SVG to get good tracking.
The issue is that with a good modern motor, you are looking at a reactive current around 6 times the motor rated current at zero speed, around 3 times at 50% speed and about 0.2 times at just below rated speed and if you want a good result, you need many steps in around 2 - 3 seconds.
I have seen it done on a very high inertia load where there is say a 30 second start up time, but a pump is so quick..... If you use DOL on a sub pump, it could be all over in half a second!!

With a soft starter, you potentially could be looking at a max KVAr of 300% and an acceleration time of closer to 3 seconds. It may take longer to get to half speed but accelerate quicker from there, so keeping the KVArs in synch with the motor slip is a challenge. Probably easier to use a harmonically mitigated VFD at that point.

Mark Empson
Advanced Motor Control Ltd

 

[Marke]

I would normally use a 4-20mA control valve on the discharge side of the pump, start with the valve closed to about 10 - 15% open and with submersible pumps, the minimum start current in this state, is commonly around 250%. When the pump is up to full speed, open the valve slowly in a manner to limit the maximum motor torque to around 95% with a dead band of around 5% using a stepping algorithm rather than a PID.
This allows for a long empty pipe and a flat curve pump and starts the pump at minimum current, and then fills the line without pumping off the end of the curve, or overloading the motor.
Gets the pump to full speed very quickly, and the pumps operates always with full back pressure, so no thrust washer wear. Pumps last for years!!

Mark Empson
Advanced Motor Control Ltd

 

[waross]

Thanks for the information Marke.
Yours
Bill

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[Loghman_ra]

Thanks Mark ,
Good tips,
but I don't know if it is allowed to use such solution (controlled valve on discharge) in cases where ESP are used for emergency case like evacuating the floods, etc ?