Sorry for the long post. Good discussion...
From a maintenance point of view even double ended gear preferably MTTM especially with arc flash considerations has considerable advantages because you can actually isolate any particular section even to do bus maintenance.
"Considerable advantages" is what I am debating here. Again, with the shutters closed, there is virtually no arc-flash hazard and using RELT or having bus-differential significantly reduces the hazard. Of course this is predicated on the assumption there is no reason to service the 4" of line- and load-side bus stubs in the tie section. Also, as mentioned, if the customer is already willing to drop half the bus load, a MTM arrangement works just fine. Don't get me wrong, if the customer was still insistent on getting the MTTM, I would obviously just sell it to them.
The extra tie addresses maintenance on the tie cell itself.... I rarely see it so the tie suffers and becomes very unreliable over time.
Realistically, you can still service the tie in MTM applications. Simply rack it out and service it. There is no reason why it would become unreliable over time unless it was intentionally neglected. As discussed, the only thing you cant service with a MTM arrangement is the tie line- and load-side bus stubs, which is not a typical thing to service anyway. I refer you to manufacturer's publish data, NFPA 70B and the ANSI/NETA MTS standard. Servicing the 4" of bus stubs is not something one typically does or generally cares about (however, I know some people really like wasting their time lubing bus stubs).
At the distribution level unless you are running 69 kV (SF6 territory) modern breaker speeds are 3 cycles with a potential for 1 cycle relaying plus at least a half cycle to recognize a fault assuming the designer doesn’t get stupid with lockout relays and isolation relays going to the trip coil so 4.5 cycles is a reasonable minimum opening time.
No legitimate design relies on a lockout relay alone to trip the breaker. The correct way to trip using a lockout relay scheme is with a direct trip in addition to a parallel lock-out relay
re-trip (which will also blocks the closing circuit).
Due to uncertainties between relays with induction disk they used to recommend 0.35 s minimum between switching layers so between the final load device and the first layer even with the fastest devices you were already up against the 30 cycle withstand rating. With modern devices 0.1 second delays between trip curves is not unreasonable.
Just a side note for those who don't know: According to the IEEE Buff book, the minimum recommended Coordination Time Intervals (CTI's) actually depend on the protective devices you are trying to coordinate and whether or not they are calibrated. This can range from simply maintaining clear space or having 0.3s of separation between TCC's at the prospective fault current.
Partial bus differential tripping isn’t NEC required and you are assuming the tie is a breaker complete with its own over current protection. In many systems the tie is completely manual. It may be just a disconnect switch. I’ve even seen “wired” (bolt together) ties with no gear at all. Utilities often speak of emergency ties here in hurricane prone areas. You still have bus protection in both buses but the protection zone is now larger, both buses are combined. If you do have protection at the tie then you can use the tie to trip on a bus fault on the second bus either by slowing down the mains (usually not a good idea), ZSI (not so bad here within the same switchgear), or partial bus differential relaying. It’s an option, not required. There are other options.
Refer to NEC section 240.87(B). You will see ZSI, bus differential, and Energy-Reducing Maintenance Switching (RELT - GE term) listed as the acceptable methods to reduce clearing time. Also, yes, I realize some tie breakers are just really expensive switches, but this is not the best design choice if you cannot afford dropping the entire bus load while in the single-ended/emergency configuration. Obviously, sectionalizing the buses can minimize downtime and aid with troubleshooting - which may be a small price to pay in the long run. Another benefit is using RELT on the tie instead of the main while in the single-ended/emergency configuration when switching or doing anything on the far end of the gear.
As to ZSI I do switchgear maintenance as a contractor. Power plants, high availability and security facilities, large industrials with cogens, you name it. I have seen abandoned ZSI but no working ZSI. It’s like a pilot wire system...great as long as the manufacturer doesn’t obsolete it and parts are available and reliable. If all manufacturers could agree to operate ZSI on say dry contact relaying with no more than 125 VDC and some kind of maximum current (relay) spec with resettable electronic fuses to protect the upstream relay we could implement ZSI universally so that we aren’t brand/model locked and can maintain ZSI moving forward. That doesn’t exist so we can’t do it.
I disagree. You can implement a fail-safe local or remote ZSI scheme with microprocessor based protective relays. Programming the discrete I/O using dry contact and 125VDC wetting voltage is completely possible. Writing ZSI logic is not difficult given all the flexibility, reliability and support available with modern protective relays.
As far as ZSI being abandoned everywhere you go, consider that most people do not understand ZSI, and while customer equipment may support this function, it generally does not get commissioned properly or even installed to begin with. So, this is an entirely different issue alltogether.
As to your other terms. Ok maintenance switches are becoming common. But it’s a retrofit solution. Most of the time you can design out arc flash down to under 8 cal/cm2 in a new system without one and there are the issues of training and understanding what it is for. And it’s not new. Induction disk relays have independent 50 and 51 functions and most microprocessor relays have at least 2-3 setting groups that you can toggle for various specialized situations. It’s just the latest fad. Don’t get me wrong I agree with the purpose when it makes sense to use it but it’s like having a synchroscope meter on a generator for manual closed transitions...great in the hands of an expert, dangerous in any other case.
The lack of a qualified work-force who is not capable of understanding or being trained on how the gear functions is a bigger problem and not a very good reason to dumb down the system protection or avoid getting with the times.
Grounding in multiple source systems is a critical design element. It can be very simple or it can be very complicated if done poorly. It is also often blamed for a variety of other unrelated issues. I think only GE actually uses the acronym MSGF, had to look that one up.
You are correct about MSGF being a GE term, but the MSGF protection concept is not one specific to GE equipment.
Anyone familiar with multiple paralleling generators knows this one very well. The problem is not limited to MTM and in fact it’s arguably worse with paralleling gensets with designs that don’t consider grounding ahead of time. I do not personally believe that relaying is the answer. Fixing the design and installation is the answer every time. But even on a radial feed system grounding is frequently an afterthought. Personally I’m a huge proponent of high resistance grounding from 480 all the way to 10 kV and low resistance (400 A) at least to 35 kV. It’s cheap, it’s lower maintenance cost, it’s selective, and let’s face it solidly grounded systems are just sweeping the problems under the rug and ungrounded systems are taking a step backward. Multi source systems just force this issue out into the open...or cause utter chaos.
Overall, I agree with what you are saying here. As far as I am aware, relaying is the only solution for multiple parallel sources supplying a 4w system with unbalanced loads (I should have clarified that).
