Power lines connecting to the ground

[DrClaude]

This is going to be a very basic question, but something that has bothered me for a while. It was triggered again by this thread from @OmCheeto:
https://www.physicsforums.com/threads/new-deccan-trap-turns-out-to-be-downed-power-line.970313/
Why is it that when a power line is downed, there is no circuit breaker somewhere going off? Is it that, even with a direct connection to the ground, the current is still within the usual range?

 

[davenn]

Is it that, even with a direct connection to the ground, the current is still within the usual range?

That would be my initial assumption

On our local 240VAC, the pole fuse for the individual house feedline is around 50A
Now consider the rating of a breaker that would feed a whole street ( offhand, I don't know what
value it would be, going to have to ask the next linesman I talk to)

But for example, say a street with 25 homes and for evening dinner time and the avg oven drawing 25A
( usually a 30-35A CCT Breaker at the house power panel) and 1/2 of those 25 homes using their oven

12 x 25A = 300A

there could be another 5 to 10A per home for either summer A/C or winter heating per home

split the difference

25 homes at 7.5A = 187.5A

Winter time, add lighting in each home say 5A per home

25 x 5A = 125A

All the other stuff running in homes ... TV, computer, phone chargers, etc

Say another 5A per home

25 x 5A = 125A

Now haven't added stuff like hot water heating and other intermittent stuff
dishwashers, clothes washers, clothes driers etc

Just at a bare minimum of ~ 740A for a 25 house street

Wouldn't be surprised that the breaker is around 1000A for that street
would take a lot of solid grounding to hit 1000AA couple of years back the dropper cable from the pole to the house of our neighbours home failed
My early morning attention was drawn by the sound of arcing and the flickering blue light...

Went to investigate and at that initial stage, the rain water was running down the cable to where
the cable was damaged and it was arcing out between the phase and neutral. I called the authorities
and by the time they finally got there, the cable had fully broken and was now arcing out to the wet
concrete footpath. I stayed nearby till the power Co arrived so that I could warn pedestrians of the
danger.
So even that arching between live and neutral and then to ground still wasn't enough to blow the
pole fuse for that house dropper lineDave

 

[anorlunda]

Why is it that when a power line is downed, there is no circuit breaker somewhere going off? Is it that, even with a direct connection to the ground, the current is still within the usual range?

Yes, that's correct, at least some of the time. A short circuit has an impedance, and the lines between the source and the short also have an impedance. Therefore there is a range of magnitudes of short circuit currents, some of which overlap with the ranges of normal current.

Also bear in mind that with 3 phase circuits, there are several ways to make a short. Line-to-line short, line-to-ground, double-line-to-ground, three-lines-to-ground, and line-to-line-to-line shorts. Each type causes different levels of short circuit current. The relays look at the magnitude of currents, and at the unbalances in the 3 phases, and at time.

We call it the "art and science" of protective relaying, because both words apply to actual practice.

Time also comes into play. There are cascades of protective relays, breakers, and fuses. Hopefully, the relay closest to the short opens. If it fails to open, another relay more distant backs that up, but only after a time delay to give the first one time to act and be detected. If the 2nd level fails there is a 3rd level after a longer delay and so on. If there is a short near your house, the first level hopefully interrups only your street. The 2nd level might shut off the neighborhood, and so on. More delay and wider impact at each level.

So, when you see videos of short circuits arcing, it is common to see it continue for a long time, but not an infinite time. The first level should stop it in a fraction of a second. The highest level might take 30-60 seconds to act.

The most frequent types of short involve a single line to a tree branch or a squirrel. Those are usually self-healing as the tip of the branch burns away or the squirrel falls. We call that fault clearing. The impact on customers is less if the fault clears itself in a few seconds, as compared to sitting in the dark waiting for a line crew and a truck to show up and reclose the breakers manually. Therefore, we take some risks. Some short circuits are allowed to persist hoping that they will self clear. The breakers trip after a time delay when the logic decides that it will not self clear.

There is also logic that attempts to re-route the power to customers, by opening some breakers and closing others. Sometimes, that works, but sometimes it just results in feeding the short circuit arc from the opposite direction.

We also use automatic reclosers. that may close the breaker once again several seconds later, in the hope that the fault self-cleared. If it did not, the recloser restarts the arc. Automatic reclosers may try, 5 or even 10 times before giving up, so the arcs at the fault location may be seen to start and stop repeatedly over several minutes.

You may see the net result of these various schemes at home. The lights dim, then go out, then they may flash on for a moment, flash again, and finally come back steadily after 30 seconds or a minute. You see the net result of several layers of protective strategies and self-clearing behavior of the fault. Customers a few streets down may experience a different sequence than you.

I can't resist a plug for my personal hero. Much of the basic methods and strategies for protective relaying were invented by Thomas Edison as an integral part of the Edison Electric Illuminating Company. Edison did much more than invent the light bulb. He conceived the whole industry and power grid including all necessary attachments and accessories all the way down to insulators and brackets to hold the wires, operating procedures, even billing methods. He delivered a turnkey electric utility on the first instance in NYC. Then he delivered a complete second instance for the Paris Exposition of 1889.

Thanks @DrClaude for giving me the excuse to expound.

 

[OmCheeto]

Why is it that when a power line is downed, there is no circuit breaker somewhere going off? Is it that, even with a direct connection to the ground, the current is still within the usual range?

Well, if you've seen my latest post in that thread, you'll see that I calculated that only 1% of the capacity of that particular line was required to melt 2800 kg of basalt in only 86 seconds.

Without knowing the particulars of the voltage and capacity of the line in India, we'll probably never know how long the line was down.

On a somewhat totally unrelated note...
When I finally found a site that listed the resistivity of basalt, they listed a range of 1,000 Ωm to 1,000,000 Ωm.
Taking the average of the two, as I didn't feel like doing the maths twice, I came up with 500,500 Ωm. That seemed way to far away from 1,000 Ωm, so I averaged the exponents, and came up with 32,000 Ωm.

Is there a name for such exponential averaging?

 

[nsaspook]

Eventually the current draw will trip the line breaker but beware, it double taps (automatic reclosers).



NSFW language

 

[nsaspook]

This one is my personal favorite for power line faults because of the duration and wide area of continuous faults.

 

[DaveE]

Is there a name for such exponential averaging?

Geometric mean

 

[sophiecentaur]

There is plenty of scope for much better protection these days with The Internet of Things. Simple Voltage sensors, placed at intervals along a line could identify and locate a sudden load applied in an unplanned / unserved location. Any large load could be required to announce its presence in advance to the system. The possibilities are limitless and I'm sure it's being worked on.

 

[anorlunda]

There is plenty of scope for much better protection these days with The Internet of Things. Simple Voltage sensors, placed at intervals along a line could identify and locate a sudden load applied in an unplanned / unserved location. Any large load could be required to announce its presence in advance to the system. The possibilities are limitless and I'm sure it's being worked on.

You're quite right Sophie. I used to daydream about such things myself. In general, applying local intelligence to control and monitoring functions.

But then came the cybersecurity demon. Any device that communicates with the Internet is a potential threat vector. Distributed IOT devices would be a nightmare. Even innocuous sounding smart meters at houses are hugely controversial. They started a new branch of conspiracy theories.

What you need is a device that gathers information locally, and acts locally, yet having a global effect. Something that can't be remotely hacked. Christian Huygens invented such a device in the 17th century, the flyball governor. Attach one of those to every generator, and you achieve global frequency control of the grid.

It is called the Battlestar Galactica defense. You can have computers, but you can not have a communications network because networks can be hacked. There really are some new devices for power distribution that fit that description.

 

[sophiecentaur]

Any device that communicates with the Internet is a potential threat vector.

You could use WhatsApp - that's pretty secure or there are point to point systems that could use the power lines themselves - too accessible, perhaps?

 

[anorlunda]

there are point to point systems that could use the power lines themselves - too accessible, perhaps?

We can and do use non-Internet communications. For the most critical measurements we use analog signals over dedicated copper wires spanning hundreds of miles. They are routed to be geographically distributed, so that a reckless crew with a backhoe can only destroy one wire at a time. But that is outrageously expensive and used for only a few things.

Power line carrier is insecure and very low bandwidth.

Microwave line-of-sight is pretty good, but expensive. You could not practically have one microwave (or laser) transmitter every 100m over a long power lines.

I once wrote up an idea using infrared LEDs on the power lines, sensing current and powering themselves with a current transformer. We could put one on every pole. A tethered blimp in the sky would detect those IR signals for an entire city, or a satellite for an entire continent. It is a fun daydream, but hardly practical.

Military battlefield communications technology could point the way. But your good idea for local smart devices, implies tens of thousands of them, maybe a million instances. Affordability dominates. It would be analogous to the army having each bullet communicating with the Pentagon while in flight.

 

[OmCheeto]

Anyone know if they have any automatic "GFCI" type technology designed for power grids? I googled a bit yesterday, and couldn't find any evidence.

This one is my personal favorite for power line faults because of the duration and wide area of continuous faults.

That was really fascinating, but the comments section didn't tell me what I wanted to know, ie, "What the hell is causing that?".
Most of the guesses looked like man-splaining; "...All these failing in a cascade like that..."

I found an identical, lower resolution video that the OP [youtube, not this thread] had posted a week earlier, where he made the comment; "This is the aftermath of a pretty brutal thunderstorm in Fort Worth Texas on May 10, 2011. It was taken from my balcony on the 34th floor of a building in Fort Worth. Though I thought we were at war or was terrorism, it was a massive series of downed 12,470 volt power lines.".

For some reason, I couldn't understand how a thunderstorm could cause such a thing, and I was also suspicious of the "downed power lines" explanation. So I looked some more, and found a few articles.

​

NPR​

The local electric company tells Huffington that what you're seeing are flashes created when lightning hit "feeder" lines that carry electricity to neighborhoods and the power surged through transformers and other equipment.​

​

​

CBS News​

A lightning storm caused several transformers to burst...​

​

"[National Weather Service meteorologist Matt] Mosier said he didn't know if it had to strike the transformer directly or the ground close to it but if the lightning carries a large enough electrical charge, it can "blow" a transformer."​

​

​

Huffpo​

The storm left about 4,000 people without power,​

Matt Mosier, a meteorologist with the National Weather Service, told the Fort Worth Star-Telegram that there were more than 120 lightning strikes from 9:00 to 10:00 p.m.​

​

Anyways, lightning ground strikes are relatively rare where I live, so those 120 lightning strikes to power line poles, and/or the lines themselves, overloading the transformers made sense to me as a reasonable cause.

Amazing that only 4000 people were without power. The Dallas–Fort Worth metroplex has over 5 million residents. Sounds like they have a well designed, robust local grid.

Anyways... Getting back to the OP [this thread], I viewed the original problem as being somewhere between fulgurite creation, and electric wood burning:



Which is why I decided to melt a cube of basalt, rather than go through doing the maths on what actually happens, and end up sounding like a total mansplainer.

I would imagine that if the Indians dig up their now extinct notavolcano, they will find something that looks very much like a combination of both.

 

[anorlunda]

Anyone know if they have any automatic "GFCI" type technology designed for power grids? I googled a bit yesterday, and couldn't find any evidence.

Sure. 3 phase unbalances. But the nomenclature is zero sequence current, not GFCI.

 

[Tom.G]

Here in California, the electric companies are in trouble for starting huge wildfires. One of them, San Diego Gas and Electric (SDGE), states they have configured, and added to, their existing substation instrumentation to sense a broken HV transmission line and remove power from it before the wire hits the ground.

Seems like it will help. We'll see.

 

[anorlunda]

San Diego Gas and Electric (SDGE), states they have configured, and added to, their existing substation instrumentation to sense a broken HV transmission line and remove power from it before the wire hits the ground.

Yes, that could work. But it might be vulnerable to false positives resulting in lower reliability.

Trees touching wires triggered the massive NE blackout in 2003. https://en.wikipedia.org/wiki/2003_blackout

The California fire problem is serious and difficult. If I was the PG&E risk manager, I would not depend on technology to protect my company from liability. I would order the whole grid to be shut down in all counties where and when the fire risk is red. Of course that would get me tarred and feathered and run out of town on a rail.

They could go entirely underground, but that is hugely expensive and it would take a decade to complete. Underground brings its own risks. In 1998 Aukland, NZ was blacked out for 5 weeks because hot weather overheated all the underground cables at the same time. That was a classic example of common mode failures.

 

[Windadct]

Also - phase to phase fault, like a tree branch can still cause a fire, and not be a ground fault. A GFCI - is a complete device that senses ( very low leakage) and interrupts the circuit. Just adding Ground Fault, current imbalance or zero-sequence protection will not catch all of these, I would even argue it would only catch about half of tree branch type issues. {I'd rather not bring up the whole de-reg of the utilities issues, but placing profit ahead of both safety and reliability caused many changes in the way Utilities operate, the effort going into right-of way maintenance ( cutting trees back ) is an issue, and then local authorities(townships / munies) started suing the utilities to PREVENT cutting back the trees ... - this alone is not a trivial factor}

The basic concept of protective relaying is straightforward, but the actual variety of situations and protection methods employed can be quite complex, thus the need for dedicated protection engineers, it is its own field of study.

At the distribution level ( 2400 to 36KV) overhead lines may only be protected by fuses which only protect against overload / high current faults. Adding something like a re-closer (which can be tripped offline by a more complex relaying scheme, or remotely) at every circuit branch would become very expensive, and add to the coordination complexity - especially if considering areas where multiple feeds are possible.

Still the relaying IS becoming more complex or "intelligent" - just a AFCI ( Arc Fault Circuit Interrupter) in a home "looks" at the current waveform to try to determine if there is arcing in the circuit, there are relaying system that are much better that determining abnormal conditions, not just current levels and balances ( but cause nuisance trips on dimmers and treadmills). But they again all add to the complexity, require more engineering and monitoring, and higher skilled people in the field setting them up.

Lastly - tripping off power, just to prevent a fire, still present many other risks. Traffic lights are a good example, even if they have aux power the transition from utility power to local ( blinking red/yellow) still is quite likely to cause accidents. People that are shut in - may need the AC and are not able to deal with the heat, and these would be the days where the fire danger ( and smog) is high. Liability is not just about the fire danger - it can also get hit by the productivity downtime of the businesses being fed, ever shut down a refinery? ( fun!)

 

[Fisherman199]

Fault impedance can vary from nearly zero to effectively infinite (open circuit). When engineers examine fault characteristics of a power system one set of figures will represent the fault duty of the system itself (the fault level the system will provide given the lowest impedance path back to the station transformer). Another set of numbers will be given to represent some fault with an impedance value. 30 ohms is common for rural areas. 10 or 15 for more densely populated regions. This is more art than science.

I'm not sure how common place high-impedance faults are but their existence is one of nuisance. High-impedance faults have a current level comparable to that of normal system parameters. They're impossible to detect using only TOC. Difficult but not impossible to detect with directional consideration.

There exist relaying schemes not widely used that claim to provide effective detection of high-impedance faults but I'm not sold on them just yet.