Paralleling MV Generators 1

[Mike6061]

We have MV generators that distribute at 4160V.For plant operations the generators are connected to a common bus. Up to (4) feeders originate from this common bus through SF6 switches that provide overcurrent protection. The generators have a nuetral tap that can be bonded to ground but typically we disconnect the bonding jumper and ground the chassis and distribute using an ungrounded system. I just ran into a situation in which some newly received generators still have the bonding jumper in place. We want to parallel these generators with the ones that aren't bonded. The obvious solution is to disconnect the bonding jumpers on the new generators and run the plant ungrounded. But if I didn't remove these bonds what problems would this create if any? Would the whole plant simply be seen as solidly grounded or is there some other consequence to paralleling ungrounded and grounded generators? Would there be a difference if each generator was grounded individually versus on a common grid?

Thanks in advance

 

[dpc]

If you have one generator on-line that is grounded, the entire system is grounded. The other generators can be left ungrounded but the system is still grounded.

For your consideration - for industrial-type systems:

Ungrounded 4160 V systems are not recommended.
Solidly-grounded 4160 V systems are not recommended.

Normally, a system such as you are describing would use low-resistance grounding through one or more grounding resistors to limit the ground fault current while still maintaining a good neutral reference point.

You might want to do a little research on MV grounding.

 

[YuriSarcas]

Every MV generator system I have done has a Neutral grounding resistor for each generator and this usually goes through a disconnect switch of some sort.

I would think that if you have one generator grounded and the rest are not and you have ground fault protection it might give you trips on big transformer in-rush loads.

 

[Mike6061]

Our whole system is known as the Deployable Power Generation and Distribution system (DPGDS). It is used for primarily for military applications. The generator units are rated at 920KW. Typically we will parallel up to 4 of these generators (depending on the load). The units come with a Primary Switching Center with ground fualt protection along with 150KVA XFMRs with feed thru capability. The generators are equiped with voltage regulators, load sharing module, overcurrent, undercurrent, over voltage, undervoltage (every relay imaginable). It's a plug and play system that is operated remotely. All connections are facilitated using load breaks. Typically distribution is either underground or surface laid depending on safety considerations and duration of the event. We have considered HRGs or LRGs but can't seem to justify their use. One way or another a ground fualt on our system will facilitate the operation of a protective relay. I am waiting for somebody to make a very convincing argument that in my circumstance an LRG or HRG will make a significant difference.

 

[sberbece]

Mike6061,

In a generation – distribution system the generator is the weakest link from the short circuit withstand point of view.
A metallic or bolted Earth Fault close to generator is generally higher than a 3 phase bolted fault @ generator’s terminals (For more details please see IEEE 142/2007 – Grounding or Green Book and IEEE 242/01 – Buff Book, IEEE C37.102 – AC Generator Protection).

Usually, in case of LRG the earth fault current is limited from 200 A up to 150 % of generator’s In.
In case of HRG the earth fault current is limited to 3 to 25 A. With this method, the generator winding can be protected up to 95 %.
However, the decision of whether to employ LRG or HRG for protecting a generator or a power plant having few generators working in parallel is to be taken by a qualified electrical engineer.
In few words, it shall be decided how much the generator stator winding shall be protected and how much it will cost.
It seems that for your voltage level, is common in USA to use LRG.
Regards,
SB

 

[dpc]

The main advantage of resistance grounding is to limit the damage caused by ground faults. This is especially true for generators and motors. When a stator ground fault occurs inside a generator, limiting the current greatly reduces the damage caused to the core steel. If damage is limited to the winding itself, it can be replaced fairly easily. If the core suffers significant damage, the core must be taken apart, repaired and restacked. The downtime and cost implications are significant. Especially since 70 to 80% of faults start as ground faults.

If you don't care about the generators, transformers, motors or other equipment, then consider reduction in arc-flash energy for ground faults and the reduced risk of ground faults escalating into three-phase faults.

All relays in the world won't take the place of limiting the fault current.

 

[Mike6061]

dpc

What you had to say makes alot of sense to me. A big reason most of my collegues are reluctant to institute LRGs is simply becuase we have never experienced a problem running an ungrounded system. However, eventually a ground fault will occur. I think I could sell the LRG on the basis of equipment protection and personal safety. I will contact the manufacturer and ask for their input and proceed from there.

Thanks

 

[waross]

"I think I could sell the LRG on the basis of equipment protection and personal safety. I will contact the manufacturer and ask for their input and proceed from there."

Probably a typo. Did you mean High Resistance Grounding? This is the system with the best equipment protection.
Low Resistance Grounding can still do a lot of damage to a stator or other equipment.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[dpc]

Bill is correct, High resistance grounding is best for limiting equipment damage and arc-flash.

The main disadvantage to high-resistance grounding is the lack of selectivity. With multiple generators on a common bus, a ground fault anywhere on the system will shutdown the entire system. Assuming you elect to trip. It is possible to continue running with a single ground fault with high-resistance grounding, but I'd strongly recommend against it, especially if this is a military application.

Low resistance grounding allows for selective coordination. It is a tradeoff since equipment damage will be greater.

Traditionally, LRG systems had 400A or even 1000A resistors. There's no reason for that these days. Selective coordination can be achieved with 100 A or 50 A resistors in most cases.

High-resistance grounding would limit current to less than 5 A.

Either approach is a big improvement over solid-grounding or ungrounded.

 

[waross]

Another comment in the form of a question:
In Central America, Caterpillar does not rewind generator ends up to over 1.2 MW. In the event of a burnout, the generator end will be replaced with a new end. Under these conditions, protecting a stator from arcing damage is pointless.
The question, does anyone know if this is Cat policy world wide or is this a local dealer policy in one or more Central American countries.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[catserveng]

Back when I was working with CAT, it was more of a Latin American District policy, mainly based on the fact that we had a very hard time finding shops to do competent rewinds and repairs to larger generators, from a cost and time standpoint, plus trying to prevent repeat failures, replacement with new was often used. EAME, the Europe, Middle East and Africa District also applied a similar policy in many areas of the world.

CAT Power Generation had a hard time in dealing with this particular issue, we put a lot of modular and permanent plants in that in my opinion we not properly grounded or protected, but went into service anyway, and we had a fair number of failures. In older projects with more physically robust generators we had fewer problems, in newer plants with "value engineered" generators we experienced more problems and tail end failures. We ate lots of diodes, killed lots of AVR's and I pulled a lot of tail ends, not sure exactly where the cost saving was.

Did learn that a well thought out protection scheme, properly tuned and commissioned control systems, and good engineering practises helped prevent failures, too bad too many times the cost of doing it right the first time prevented things from working the way it should. Of course many of our competitors had similar issues as the drive to reduce cost caught a lot of us.

I did a lot of modular and portable multiple units systems all over the world, mostly 480V and 12.47kV for CAT, we usually opted not to use grounding resistors or reactors on systems with less than four units, at four units and above the amount of available fault current was more than enough to kill a stator. Of course we did have two and three unit systems that had stator failures, so this wasn't a perfect solution.

To Mike6061, I know one of the criteria for the military mobile systems is to minimze cube and weight for transport, and generally the deployment of these systems was supposed to be "short duration", but as I'm sure you've seen, many of the plants go into service for longer periods of time, and are exposed to more potential risk for faults, and don't get the PM maintenance they should while in service in those conditions. The application of proper grounding protection would limit potential of catastrophic failure in the event of a ground fault, so far you've been lucky, but it will eventaully catch up, and Col Murphy has a lousy sense of humor and timing.

Lot of smart people above have given you some pretty helpful information, hope that helps.