I have a customer with several large service installations. 3000 to 5000 amp
600/347 volt. Most of these low voltage switchboards are close coupled to
the transformer with buss duct. calculations indicate that incident levels are above cat 4 at the main secondary device (line side)
I have the following questions:
1- Do we classify this main device based on Line side AF levels or loadside
2- Would better primary protection help to lower these AF levels at present
we have S&C SM 40 fuses.
3- Are then any products out there that could be installed at the transformer secondary bushings that would lower AF levels.

1. Line
2. Maybe, you have to play around with different fuse types in your software
3. With a switch on the primary this is hard, one solution is to put a mini vacuum breaker where the fuses are and use differential or arc flash relays, but it is expensive. Most people use remote racking and switching for the main LV breaker as a solution to this very common problem.

Transformer Secondary Blind Spot

Per the recently released "Complete Guide to Arc Flash Hazard Calculation Studies", Jim Phillips, pp 122 and 123, the secondary side of a power transformer is typically a "Blind Spot" for arc flash mitigation.

I have worked with a transformer supplier, located (curiously enough) in Guelph, Ontario. For custom transformers, Overcurrent Protective Devices are built right onto the transformer secondary buses (multiple secondary winding transformer).

The OCPD line side is connected to xformer secondary bus, we wire to OCPD load side. This is only helpful for new installations, but it is a viable solution. Protects area between transformer and first load panel as well as adding manual isolation capability. Does not protect against arc flash event within the transformer.

For you familiar with Southern Ontario, the transformer builder is located a stone's throw from the well regarded Sleeman Brewery...

AB

Nice plug

interesting, any links or info to post?

Ok, eh, I found this mouse, in a bottle of your beer

The OCPD at the transformer bushings could work for a new installation and It is something that transformer manufactures can offer, Our problem is with the existing installations, there isnt the space for this option. In most cases we are using oil filled TLO units that are directly coupled to the 600 volt swbd located inside the building. I am presently looking into better primary protection to try and lower these AF levels.

What exactly do you mean by (blind spot) is this an area that is ignored? Due to the difficulty in solving the problem? Such as in the case of a unit type substation.

On primary side of the subject power xformers, possibly put a vacuum contactor in series with primary fuses. This can be a control point for a relay protection scheme.

Toshiba or Joslyn contactor or similar is physically small relative to an entire MV CB. Check your local codes for suitability.

As far as "blind spot", please refer to the book I referenced. Jim's book is very helpful.

What ever device you put there must be rated to interrupt the fault current without relying on fuses. Many of the 'motor starting' contactors are not rated for this type of 'stand-alone' operation.

OK I will buy the book!
At present the solution is to install breakers at the TX secondary. For new installs its not a big deal. For an existing service its a lot more difficult.
I will investigate the primary protection route further. tks

The "Blind Spot" is a term that we used back in my electric utility days. It is simply the location between a transformer secondary bushings and the first secondary protective device.

In other words, it is the area on the transformer secondary that is only protected by a transformer's primary device.

Nice thread, I'd like to hear more about these "blind spot" solutions. This is seems to be a very common problem. It's suprising that there is so few published solutions.

Jim's "Blind Spot" is a great name for the transformer LV connection because it is always difficult to protect. Sounds a bit like "turning a blind eye" or ignoring it because from a risk point of view, if the LV tails are short and nobody opens up the cabinets then this should be low. There is also a term that is used to describe the droppers from busbars onto protective devices or motor control units which is the "no fault zone" (I don't know if this term is used in the US and Canada). The lower likelihood of a fault occuring on these cables is a justification for having very much reduced conductor sizes. In my experience however faults do occur in transformer connetcions and in "no fault zones" and usually with serious consequences.

Smaller fuse sizes within the same line do trip faster IN GENERAL. Faster trips = lower AF rating. Faster fuses also achieve the same thing. Quite often just sizing a fuse closer will do it. Even from S&C there are lots of options.

Be aware that this does not always work. If you are dealing with current limiting fuses, the current limiting function can sometimes actually extend trip times and fault energies and make the problem worse.

Prior to even worrying about arc flash the typical way that fuses were sized was to look at the damage curve of say the cable itself. This resulted in very long trip times. A second pass might be done to look at say the transformer impedance only (cable impedance = 0) and then adjust the fuse down low enough so that it at least trips on this.

This approach satisfies protection requirements (SCCR) but nothing else.

With arc flash as a concern reducing the fault time and current limit is more important. So you tend to look for the arcing fault current (85% or 100% of bolted fault INCLUDING the cable impedances) which is the actual fault current anyways and set it to this.

At the secondary side until you hit an overcurrent protective device, the only protection is on the primary side. On the primary side of the transformer the fault current appears after being transformed down by the transformer.

Using traditional methods then, usually the primary side device is limited by the transformer inrush current so that you simply can't set it tight enough to achieve much protection at all.

If the primary side is a circuit breaker with an external trip unit, there is often a relatively simply solution. If you install bushing CT's on the secondary side and connect them to a trip unit, you can trip the primary circuit breaker and reduce the size of the "blind spot" to effectively nothing.

However frequently the best/most cost effective protection on the primary side especially in substations tends to be fuses, which as I said, just can't get there very easily.

The other big problem with the "blind spot" is that this is the point where the available fault current is at the greatest value. And if you are interested in pushing efficiency in the transformer (reducing core losses specifically), this pushes the transformer impedance down rather than up. Otherwise, purposely setting the transformer impedance fairly high is an effective way to reduce the available fault current dramatically.

Another technique?

In another thread, the topic of "Current Limiters" was raised by a member. These are special inline fuses that are attached to each conductor.

Like this:

Xformer bus === CL fuse ~ ----- conductor ----- [ ] next OCPD

Each parallel conductor at the xformer secondary would have a current limiter fuse.

I am not experienced with these current limiters, and I'm not sure how to model them in SKM. The limiters are in the default SKM library, and appear in the TCC models.

Per the current limiters datasheets I located devices are UL listed (2 of 3 vendors) but do not list cUL or CSA.

I'm curious about how they would reduce incident energy. If you have many cables per phase and one limiter per cable (not per phase) do you assume they all interrupt at the same time if there is a downstream fault?

I made a quick SKM model including the current limiter fuses. Did arc-flash study with and without current limiters. The current limiters, when present, had little or no effect on the arc flash result. It was just a quick look, I might have missed something. Comments?

I don't know if these are 'special' current limiting fuses, do they differ from regular (CL) fuses?

As for your model, fuses that are sized to the transformer nominal current will have bad arc flash performance. The current limiting effect will typically start at 15x-20x nominal current, and for low voltage systems your arcing current will simply not reach that value. It will eventually trip on it's thermal part of the curve, but that won't help you much. CL fuses are not a panacea, they'll only reduce hazards if arcing currents are larger than the fuse's CL cutoff.

Nothing special about the fuses. It does not eliminate the incident energy inside the transformer enclosure. It only helps on the cable and around the OCPD (circuit breaker?) located downstream of it. The problem that I typically run into is the incoming main circuit breaker in a panelboard which makes the whole panelboard prohibitively high, or else for convenience to lower the main circuit breaker rating. If you use the cable limiter fuses the way they are intended to be used they will have little or no effect because they are designed around the cable thermal limit, not the arcing fault current.

If I had a circuit breaker on the primary side and it was not physically too far away, and had a shunt trip or better yet, a separate protection relay, another strategy would be to add CT's to the secondary side of the circuit breaker and tripping the upstream breaker. If money was not a big issue, adding PT's to both sides allows for full blown 87 relaying but normally this is cost prohibitive.

Lately I've been mostly working with metal enclosed gear. I'm becoming quickly less enamored all the time with metal clad gear that has failure rates an order of magnitude higher than metal enclosed gear. Even in a mining application (where I do a lot of work), the failure rates for metal enclosed gear are very low compared to the issues I've had with metal clad drawout gear.

Terminology, error by me. The products are called "cable limiters" or "cable protectors". They are fuses with bolt or crimp lugs to allow connection to cable or busbars. No fuse holder needed.

Thanks for the replies.