NYC power outage

infinity

Moderator
Staff member
The building I'm currently working is in that zone of black out. Wonder if the generator started. :jawdrop:
 

gar

Senior Member
190714-1945 EDT

The reporting on this problem is interesting.

It has been said that about 70,000 customers lost power. I am not in the power industry and have little knowledge about it. But searching for a few tidbits and knowing what my power consumption is I can make some assumptions.

During a day I am in the range of 1 to 2 kW for a load. So assume these NYC customers are at 1 kW, then they represent a load of 70,000 kW or 70 MW. How big are large distribution transformers? Possibly 100 MVA or more. These apparently take 6 months to several years to make, and many may not be made in the US. Also very expensive, and many units could not be sitting as spares. Also they very heavy and difficult to transport.

It seems that the outage occurred all at once. An all at once event, like the great eastern blackout, has to have some common point of origin. However, it might be something small that cascades to a big event.

I doubt one large transformer failed. Power was restored too quickly.

In one report mention was made that the fire department responded to transformer fires. Implies more than one transformer. Why would more than one transformer fail at about the same time, possibly at different locations?

Another report mentioned a mechanical failure, that likely does not mean electrical.

It is clear that the politicians are technically ignorant as are the reporters. Most of the reporting is of an emotional nature.

Can some of the power people comment on their thoughts.

.
 

yuhong

Member
190714-1945 EDT

The reporting on this problem is interesting.

It has been said that about 70,000 customers lost power. I am not in the power industry and have little knowledge about it. But searching for a few tidbits and knowing what my power consumption is I can make some assumptions.

During a day I am in the range of 1 to 2 kW for a load. So assume these NYC customers are at 1 kW, then they represent a load of 70,000 kW or 70 MW. How big are large distribution transformers? Possibly 100 MVA or more. These apparently take 6 months to several years to make, and many may not be made in the US. Also very expensive, and many units could not be sitting as spares. Also they very heavy and difficult to transport.

It seems that the outage occurred all at once. An all at once event, like the great eastern blackout, has to have some common point of origin. However, it might be something small that cascades to a big event.

I doubt one large transformer failed. Power was restored too quickly.

In one report mention was made that the fire department responded to transformer fires. Implies more than one transformer. Why would more than one transformer fail at about the same time, possibly at different locations?

Another report mentioned a mechanical failure, that likely does not mean electrical.

It is clear that the politicians are technically ignorant as are the reporters. Most of the reporting is of an emotional nature.

Can some of the power people comment on their thoughts.

.
I suspect it is the transformer that steps down to 208V.
 

gar

Senior Member
190714-2202 EDT

An added comment not related directly to the NYC problem.

Our shop is getting a new (replacement) substation. This is about 1/2 mile away from the shop, and about 1 mile from Michigan Stadium.

This substation is being built like a fortress, high walls and tall posts.

My son told me that he saw a trailer bring in a new transformer to the substation. This is located on S. State Street. The transformer was large, the load was oversize, and the trailer had very many axles. The road had to be closed to thru traffic. I believe that a standard semi-trailer with one or two axles can carry 40,000 #. My guess is that this transformer was well over 100,000 # (50 tons). Not something you can move and install quickly.

None of the news comments on this replacement substation indicate its power capability. We also are getting a new added large substation on the North Campus about 5 miles away. The North Campus will then be supplied by two substations to provide redundant power. The other substation is not the S. State location, nor is it the U of M Main Campus power plant.

The U of M North Campus has a number of large backup generators, and a small gas fired power plant. I don't know the power requirement of North Campus, but power has to be supplied with little interruption. A lot of power in a small area.

For a high power illustration note that at one time the Ford Rouge power plant had a capacity 345 MW. This supplied an area of about 1 x 1.5 miles.

.
 
Last edited:

Jraef

Moderator
Staff member
What usually happens is that a smaller distribution transformer blows up, but the sudden loss of load on the system creates a voltage surge that other transformer tap changers cannot react to fast enough and protective relaying starts cascading to shut down more and more systems in a concentric ring around the initial failure. But they can't re-energize all at once until they know where the initial fault took place.

PS: This is not to say there wasn't some sort of secondary problem with their relaying, there likely was/is because it's not SUPPOSED to happen this way. So the investigation will hopefully get to the bottom of that issue.
 
Last edited:

ggunn

PE (Electrical), NABCEP certified
190714-1945 EDT

The reporting on this problem is interesting.

It has been said that about 70,000 customers lost power. I am not in the power industry and have little knowledge about it. But searching for a few tidbits and knowing what my power consumption is I can make some assumptions.


.
70,000 customers sounds, on the one hand, like a lot of customers, but on the other hand, how many customers are there in NYC?
 

kwired

Electron manager
70,000 customers sounds, on the one hand, like a lot of customers, but on the other hand, how many customers are there in NYC?
Exactly. Presuming that mostly means how many metered services are out. How many small apartments are there with fairly limited load in NYC and one meter on the system? How many small businesses with somewhat limited load?

1 large high rise building might have a good chunk of that 70,000 people in it during normal "business hours" but may or may not be counted same way as other things I mentioned above.
 

gar

Senior Member
190715-2016 EDT

ggunn:

I was using the 70,000 customers to try get an estimate of the size of a single transformer to support that load. But I also don't think a single transformer was the cause. That is why I mentioned a cascading event.

I have now done a search for the weight of a 100 MVA transformer. Found this site https://www.btbplaza.com/index.php/en/component/joomd/transformer-inventory/items/view/88-100-mva-220-33-kv-09
It lists a weight of 104,900 kg or at 2.2 #/kg the weight is 230,780 # or 115 tons. Huge to quickly change.

It is possible this is the size of the transformer going to our local substation.

.

.
 

kwired

Electron manager
190715-2016 EDT

ggunn:

I was using the 70,000 customers to try get an estimate of the size of a single transformer to support that load. But I also don't think a single transformer was the cause. That is why I mentioned a cascading event.

I have now done a search for the weight of a 100 MVA transformer. Found this site https://www.btbplaza.com/index.php/en/component/joomd/transformer-inventory/items/view/88-100-mva-220-33-kv-09
It lists a weight of 104,900 kg or at 2.2 #/kg the weight is 230,780 # or 115 tons. Huge to quickly change.

It is possible this is the size of the transformer going to our local substation.


.

.
Another problem is how things tie together and what may be automatically transferred when something fails. The further upstream you go into distribution the more customers will normally be impacted (some large industrial customers can skew this at times). Initial failure that triggered things could have been at lower level, if some of the lost loads begin to transfer to working sources a cascading event like you mentioned could result in even more demands on still working sources and at some point they must open some or even all their output to prevent overloading.
 

GoldDigger

Moderator
Staff member
The most recent and most plausible description, from earlier today, from ConEd, was that a 13kV cable failed and the protective relaying system intended to isolate such a fault did not function as designed/intended, allowing the short circuit to bring down interconnected distribution equipment throughout a good part of the city. If the fault was a simple short, both ends of the cable would have to be opened to allow the rest of the system to continue normal operation. A failure of the relaying at either end could cause problems, especially if cable was used bidirectionally or in parallel with another route.
So it was not a question of discovering and fixing a fault in a limited portion of the system but rather manually isolating the system and repairing collateral damage, even when the fault location was determined.

We will see what else comes out in coming days.
 

nickelec

Senior Member
I live and work in NYC keep in mind that when they say " customer" that means a single account/meter alot of buildings in the city only have one meter , so one building could have 3-400 units in it but still be considered one customer According to edison

Sent from my SM-G965U using Tapatalk
 

kwired

Electron manager
I live and work in NYC keep in mind that when they say " customer" that means a single account/meter alot of buildings in the city only have one meter , so one building could have 3-400 units in it but still be considered one customer According to edison

Sent from my SM-G965U using Tapatalk
Kind of what I was assuming any POCO would mean when they say "customer"- how many "meters" are effected.

Knowing what size of a transformer might supply a certain number of customers - some "customers" may average under 1kW load, others may average in MW ranges, so IMO you can't really estimate such a thing this way.
 

mbrooke

Senior Member
What usually happens is that a smaller distribution transformer blows up, but the sudden loss of load on the system creates a voltage surge that other transformer tap changers cannot react to fast enough and protective relaying starts cascading to shut down more and more systems in a concentric ring around the initial failure. But they can't re-energize all at once until they know where the initial fault took place.

So what happens when a 12.47kv distribution feeder with 2,000 smaller distribution transformers trips from a tree in the line?


Way off. I'm stunned you would say something like this.
 

mbrooke

Senior Member
190715-2016 EDT

ggunn:

I was using the 70,000 customers to try get an estimate of the size of a single transformer to support that load. But I also don't think a single transformer was the cause. That is why I mentioned a cascading event.

I have now done a search for the weight of a 100 MVA transformer. Found this site https://www.btbplaza.com/index.php/en/component/joomd/transformer-inventory/items/view/88-100-mva-220-33-kv-09
It lists a weight of 104,900 kg or at 2.2 #/kg the weight is 230,780 # or 115 tons. Huge to quickly change.

It is possible this is the size of the transformer going to our local substation.

.

.



Keep in mind the media tosses the term "transformer explosion" around to describe everything electrical to which everyone follows leading to the perception thats what actually happened.



From Con Ed:

https://www.coned.com/en/about-us/media-center/news/20190715/con-edison-statement-preliminary-findings-re-west-side-power-outage


A 13.8kv feeder cable shorted out underground. It happens. Up to 16 times a year according to one official report I shall post. Normally this would trip the feeder breaker at the supply substation. Rarely do more then a few customers loose power from this type of event.

NYC uses "networks" where if one transformer or one feeder cable fails the paralleled (meshed) interconnected 208 volt secondaries keep the power flowing.


In this case the primary and secondary protection at the 65 st substation failed- the short circuit persisted on the 13.8kv cable.

This caused the West 49th st substation which supplies the 65 st substation to de-energize it and Astor (Astor is connected to the same 138kv 5 feeder cables that supply 65th.

But inadvertently 49th st also cut power to 5 other 138kv cables supplying 50th and 42nd.

So at least 4 substations were de-energized as a result.
 

mbrooke

Senior Member
Con Ed released a report after the Washington Heights network outage in 1999.


The appendix lists the typical yearly medium voltage cable failures for various networks in a given year:
 

Attachments

Top