Short term source paralleling exceeding fault rating of bus

[ScottyUK]

I’m developing an LV switchboard replacement programme and am in the spec review stage prior to going out to tender. I’ve identified what I believe is a problem with the fault rating of the original switchgear and would appreciate some thoughts.

The boards are designed as a three-section type with one incomer per section and a bus section switch between the A and B sections, and another between the B and C sections. There is no interlocking between the incomers and bus section switches giving rise to a condition where we have the potential for two or conceivably three sources feeding one bar with one or both bus section switch(es) closed, a situation which will take the fault level massively over the ACB rated breaking capacity and also exceed the rated capacity of the bus bracing. The normal operating condition is for the bus section switches to be open and in this situation the board is adequately rated.

Paralleling is necessary in order to avoid interrupting the process and is controlled procedurally. The switching process is a manual local operation. Paralleling only occurs for a short period when one source has to be taken out of service or is being returned to service. My opinion is that the bar and breakers must be rated for the maximum foreseeable fault. I think the condition where three sources are in service and both bus section switches are closed could be prevented by a 4-out-of-5 key exchange interlock, leaving the worst case as two sources with a closed bus section switch between them and the third bus section on a single source.

‘Others’ here believe that exceeding the rated fault level for the short term parallel operation is an acceptable risk because ‘that is how it has always been done’. I disagree and think there is no acceptable excuse for exceeding the equipment rating, particularly when modern gear is capable of meeting the requirement even if the old gear we are replacing, which is roughly 40 years old, couldn’t do so. I’m especially unhappy because it’s likely that I will be expected to sign off this design, and I’m going to raise a few eyebrows if I refuse to do so. I had hoped not to rock the boat too much in my new job until I’d had time to settle in but it seems I’m back to my old tricks.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[racobb]

Scotty,

Check out this thread:
thread238-249427


Alan

 

[ScottyUK]

Thanks Alan,

I should have said that I had already read that thread, I'd actually commented in it too. The main difference is that this is not an auto-transfer scheme (or at least it isn't proposed to be so) which makes the transfer much longer, I expect it will be several tens of seconds.

I'm pretty confident that I am right in arguing against the existing switching policy but I am curious if anyone employs a similar switching policy and to hear some of the reasoning used to justify it.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[davidbeach]

The only justification is a penny wise, pound foolish attempt at being cheap. Manual switching raises the probability of something else going wrong. Think about what happens to you arc flash hazard levels.

At higher voltages, I've seen schemes where under rated interrupting devices were blocked from tripping and the main tripped instead when fault current exceeded the device rating. But I doubt you can successfully block a low voltage breaker, I doubt they have a withstand rating higher than their interrupting rating, and if the bus withstand is exceeded you are still in trouble.

Just replace the gear and do it right, like you know needs to be done.

 

[itsmoked]

If it would acceptable under automatic control perhaps automating the switching would be be far less expensive than the alternatives. Automation would also reduce the chance of an error in transfers in the first place.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[ScottyUK]

In my opinion it's not acceptable when under auto control either, but at least automation would keep people away from it if (when?) a problem occurs. David summed up my view perfectly.

I want a remote operation facility for future connection and manual control via a long pendant to keep the operator outside the flash protection boundary. In some instances this will probably mean outside the substation. The less up-close-and-personal interaction with the equipment during switching the better. I'm hoping to limit it to racking breakers on and off the bus, and the operator will be wearing a Cat 4 suit when doing so.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[FreddyNurk]

Agree completely with your perspective ScottyUK. I've seen similar arguments used, and really don't tend to agree with the 'thats how we've done it before argument'. There are updates to standards often happening, and certainly this indicates that one should always be ensuring that the installation is fit for purpose and engineered to the best practice of the day.

I suspect these sorts of situations shall become more prevalent as more places look to allow for paralleling of sources and/or provide backup generation with grid connect options. For that reason alone, its best to insist that everything is installed properly and with all operating situations in mind.

 

[BJC]

IMO
The time that something will happen is when has gone bad or after something has gone wrong and someone is trying to return the system to normal. I can imagine senarios where the system is being switched and the double fault occurs. I would bet a few beers that when a doble fault occurs there is someone nearby.

 

[Zogzog]

Scotty, we can wirelessly operate, charge, close, trip, and rack any breaker in North America, and it would be simple to develop there wireless devices for any breaker in the UK. And since it is wireless (You can see the operations on a LCD screen) no need for PPE. Sure beats replacing all the switchgear.

We also do vacuum retrofits for upgrading the MVA ratings of both LV and MV breakers.

 

[rbulsara]

If cornered into signing off, at least keep a written record (even informal) of you informing "them" of "they" assuming the possible risk.

Rafiq Bulsara

http://www.srengineersct.com

 

[ScottyUK]

Thanks guys.


Freddy,

Yes!


BJC,

'The perfect storm' scenario? Yes, they happen here too.


Zogzog,

The gear is really getting replaced because it's tired and old, and we are re-equipping for the long term future. It would almost certainly last another 10 years, but I doubt it will still be serviceable in another forty.

Do I assume (always risky) the wireless unit is a remote racking device and not actually radio-controlled switchgear? I honestly can't imagine the latter but some enterprising soul might have designed it! Is an example of it on your company's website?

The big problem with retrofits is getting a type test done. Type tests, especially for fault withstand, are hugely expensive over here and I doubt they are any cheaper in the US. For a one-off those costs are difficult to justify, although we are toying with the idea for the 11kV intake switchboard in the future because the cost of type testing may be lower than the cost of extended downtime.


Rafiq,

A cornered animal fights hardest! I agree an auditable trail is priceless if things ever did go wrong due to a flawed design. I don't really think it will ever come to that, I hope my colleagues are playing devil's advocate and making sure we explore all options before spending a large amount of money.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[ppedUK]

If you are specifying replacement boards and suitably rated equipment is available, then you know the answer to your "sign off" of your specs.

I have seen this paralleling scenario in a couple of LV installations, where the systems had been extended / modified over the years and it just wasn't an option to uprate all the switchgear because of a short term paralleling fault level "excursion", however, all of these installations had remote switching.

More recently though, I was involved with the design of a new 11kV system that had a short term paralleled "worst case" (all motors running, all Gen Units running, all GTs running, grid supply paralleled etc. etc.) fault level of around 75kA and 11kV gear just isn't available with that kind of fault rating. Paralleling was a regular necessary operation, but only for very short periods of time and fault limiting reactors were not an option in this installation, and yes we did consider IS Limiters. We (a committee decision via a Risk Assessment process) decided on remote switching (hardwired), but did not allow access into the switchrooms whilst paralleling took place (evacuation alarms were set off before switching) and didn't permit any racking operations whilst the system was paralleled. We also built in a 10sec auto trip such that if the operator forgot to open the parallel, the system automatically tripped one of the breakers.

As engineers, we will always opt to correctly specify the equipment and if equipment with a higher fault rating had been available to us, we would have purchased it, but sometimes things aren't always that simple. You seem to have an easier choice.

 

[Zogzog]

Scotty,

We have wireless racking but also wirless MCC bucket extractors, MCCB operators, load breaker switch operators, LV an MV breaker charge, close and trip. Not a ton of details on the site

http://www.remoterackingsolutions.com

but I can send you videos on anything if you want to contact me

http://www.cbsnuclear.com

- Scott

 

[davidbeach]

I'd say that any switching operation, auto or manual, puts you about 95% of the way to a fault. If switching puts you into a condition where you can't clear a fault safely, Murphy will ensure a fault sooner or later. Remote switching gets people out of the way, but won't save the gear or the building.

 

[EEJaime]

Scotty,
I remembered your comments in the referenced previous thread on this subject as it pertains to automatic transfers. I think your first comment in that thread was correct for that thread and tenfold so in the case here of an even longer parallelled time frame. I realize you want to cause as little turmoil as possible, and the added costs are a factor in specifying a higher fault withstand rating, but I feel that you already know what answer you will be comfortable with. I agree with you that there is little excuse for knowingly engineering in a possible point of failure. Discussion of how to protect personnel WHEN it fails is taking a wrongheaded approach. I feel as engineers we must do what we can to avoid a failure in the first place.

I may be overly conservative, but I feel that asking an owner to make an investment in long term reliability is something an engineer is duty bound to do. We may be over-riden by the bean counters, but I for one would not have my initials on the sign-off stamp.

JMO
EEJaime

 

[ScottyUK]

I've heard enough from enough people whose opinions I respect to reassure me that I'm doing the right thing arguing about this. I was pretty sure anyway, but I'm occasionally credited with over-engineering things 'a little' and also sometimes spending 'slightly' more of the company money than perhaps strictly necessary to design systems which don't fail ungraciously, don't need constant attention, and which usually get forgotten about because they quietly work all day, every day, all year. That's the way I was taught by a few old guys who'd forgotten more about electricity than I know, and I still think that way today.

I guess I'm just gonna rock the boat about this after all! Thanks for the comments - all appreciated.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[itsmoked]

Whoa folks. This is not a black and white, "only the BIG solution is the right solution", situation.

Running any and every corporation is a dance around risk.

It is the prerogative of a company to run risk if they deem it economically feasible, even to the point of human costs,(though that is risking it all and morally questionable). Just look at the Space Shuttle. So, the option to run undersized is a perfectly valid question.

Sometimes I'll destroy a wrench so it gets the job done as I have to get-the-job-done.

So if they are willing to run "over the line" for the bottom line, they are essentially trading up-front cost for self insurance, a claim they may have to cover eventually.

ScottyUK; Going on the only "correct solution" is the most expensive up-front one, when they have a historical track record that shows a much less expensive solution that does in-fact work, will mark you as inflexible and possibly unreasonable.

I recommend you NOT be inflexible. Give the undersized solution a fair shot. Do this. Cost out the difference as that IS something the Bean Counters CAN relate to. Do a rough cost of what a failure would cost. Building, approximate time to replace the gear on short notice, down time cost, and add in the harder to quantify human costs you saw at your last place during that catastrophe. Make sure you add in the cost of the automated equipment you'd have to use with the underrated solution to avoid human cost.

In all likelihood you will see that the right solution will in-fact be the least expensive solution. You will also have supported your company in all ways showing you're flexible and willing to look at all angles.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[erickench]

I would remove the incomers and bus section switches and have separate enclosures with two automatic transfer switches connected in series. Notice I say in series and not in parallel. This ATS configuration would be connected between the bus and the sections. This way only one source at a time would feed the bus.

Eric Kench, P.E.

 

[davidbeach]

Where I've always worked, the NEC has a requirement (110.10) that equipment interrupting ratings not be exceeded by the available fault current. It does not say that this requirement applies most of the time, it is an unqualified requirement. It does not make an exception for inconvenience or expense. In many jurisdictions this is the one provision of the NEC that is considered to be retroactively applicable; anything else that met code at the time of installation is grandfathered in against subsequent code changes or other changes outside the electrical system under consideration. A change in available fault current is the only thing that can cause an otherwise untouched system to cease to be code compliant. Scotty is correct, there is only one correct solution.

 

[ScottyUK]

Hi Keith,

In addition to what David says about over-stressing equipment, which is mirrored by UK legislation, there are two problems I would face with that approach, if I actually supported it at all: one legislative, and the a mix of legislative and company policy.

The site is a top-tier COMAH site, which is a UK term which essentially means that it is officially recognised as having the potential to cause a pretty cataclysmic accident if something goes wrong. In our case it the risk is of the mile-wide smoking hole in the ground variety because of the huge hydrocarbon inventory on site. The company response to that classification is to be very risk-averse and to spend a lot of money eliminating or minimising risks right across the board. In some ways that corporate attitude makes it easier to justify 'over-engineering', but it also makes for a challenging environment to work in where everything is scrutinised internally and is likely to be scrutinsed by government agencies too. It isn't quite as heavily regulated as the nuclear industry, but it's not so far behind. Risk isn't tolerated where it can be eliminated.

The other 'problem' is that UK health and safety law is founded on a principle of ALARP or 'As Low As Resonably Practicable', which allows a measure of judgement to be used in risk mitigation where the engineer is allowed to exercise a degree of judgement in balancing cost against risk. I don't think saving the incremental cost of doing it right over doing it 'wrong' would be an adequate defence if there was a fatality as a result of equipment failure.

And as I write this there's a third reason why the risk isn't acceptable: I'm likely to be one of the guys who will be operating this equipment, and I don't want my wife getting a phone call from the plant to say there's been an accident. I don't want the family of anyone I work with getting that phone call either, and certainly not because I didn't specify something properly. I work alongside these people and they have the right to expect that I'll do my job properly and make sure they have the best possible chance of going home safely. I can't totally eliminate the risk, but I can certainly minimise it.


----------------------------------

If we learn from our mistakes I'm getting a great education!