Big blackout. What happened? 40

[Skogsgurra]

When I got words about the big outage, I immediatley went to my puter to find out what my engineering friends in the US had to say about it. But no Eng-Tips page available. Of course I can understand that. No power - no Internet.

Power was restored piece by piece and I now find Eng-Tips up and running again. My question is still valid: What happened?

Glad to see you again!

Gunnar Englund, Sweden

 

[margaret695]

SidiropoulosM: Given the imperical data on birth rates that is available vis-à-vis the black outs in 1965 and 1977...... yep... there most probably will be more babies born 9 months from in the affect areas then average.

Margaret

 

[SidiropoulosM]

Margaret, you're right. Every time the demand is greater than the available generating capacity, the system frequency declines. That's when underfrequency load shedding comes into play....load is dropped in order to re-establish a balance between load and generation. As you say, the load shedding is not done on high priority loads, such as hospitals or other critical loads.

I am also interested in finding out whether undervoltage load shedding was available but I doubt it. Every reference I have seen with respect to UVLS had to do with implementations here in the west.

 

[Skogsgurra]

Having a quick look at the postings so far (it is night over here - the cat needs to be attended to) I can see that we are moving into a creative chaos and that is just the way it should be. Confusion is one of the most important steps on the road to knowledge.

Pennpoint has a point (tried to avoid that, couldn't) in pointing (oops!) out that people on this forum can - and should - contribute to knowledge that can be used to avoid similar situations in the future.

We can do that by increasing the confusion by adding more and more points of view. The confusion will then be processed into valuable knowledge (the processing has already started) after having been discussed thoroughly by all of us. We have a great advantage here. We can allow ourselves to admit that confusion reigns. That is nothing that a goverment or other authority official can do.

We have a lot of technical input so far. And some human-related things too. What about economic issues, and dollar-driven practices? Tell us about overloading and protection "adjustments" or whatever that is being done out there in order to squeeze some extra pennies out of the system.

 

[SidiropoulosM]

Skogsgurra, you bring up an important point. Transmission expansion has fallen far behing the growth in demand and generation capacity during the last 15-20 years. We have been trying to maximize the utilization of our existing transmission assets by postponing new transmission and by installing remedial action schemes (RAS) that take certain control actions for specific contingencies, to keep the system together. This has worked well so far but we can't stretch it any further. The main problem is that regulators are reluctant to grant rate increases to utilities. And utilities are therefore reluctant to invest in new facilities. We need more transmission but someone has to pay for it.

 

[alehman]

This link provides some interesting real-time statistics for de-regulated markets.

http://currentenergy.lbl.gov/

This DOE report describes some issues with the transmission system.

http://tis.eh.doe.gov/ntgs/reports.html

 

[Tnoack2003]

Wouldn't this have been avoidable if a power system analysis and simulation tool would have been in place to simulate such a scenario?

 

[SooryaShrestha]

After going through all the views here, I feel I should add some more. I think it will be helpful to all of our friends in Eng-Tips.
I start with the contingency condition in the system. Contingency condition is the one of the conditions of study, which allow to verify what wil be the outcome if a generating station or a line or a load is rejected from the system. In such study for small system like the one in developing countries where I live, the (n-1) [read En minus One] contingency condition is studied and implemented. This means what is the consequence if one line is out or one generating station is out etc. In more developed countries the present practice is to have (n-2) contingency condition study indicating "what will be the consequences if two units lines or generations are out". This gives very high reliability to the system. But the cost is also high and the loss due to such condition should be high enough to make such level of reliability feasible. Now the question is "is it enough for the US system to have (n-2) contingency condition?" from the perspective of the views of the consumers, political people, regulators, experts etc in short all the stakeholders of Electricity Supply System. Such study can be done by DOE or NERC or in my opinion by Eng-Tips also. In my personal view, the US system specially Eastern Grid needs (n-3) contingency condition considering the size and the expectations of the consumers. I feel that perhaps because of this situation, the system collapsed, though I can not surely say this without the proper Disturbance Record Data. I hope this will be seen in few days or week time.
Another important aspect of the transmission system in the present deregulated environment is that the lines have been used to their maximum thermal limit (possibly) due to the use of FACTS (Flexible AC Transmission Service) Devices, which help in maintaining the power transmission optimally by changing the power angle of a line. So with slight increase beyond capacity means a catastrophic to the system. This is, however, a risk taking for using a line to its maximum capacity.
The present day use of distributed generation using microturbines using gas or even fuel cell and solar could prove to be very efficient way of using system and way to better and more reliable condition, as it will unburden the existing transmission system both increasing stability margin and reducing line losses. This will also help sustain the Islanded system as each Island will have enough generation to maintain the system.
Another Important aspect of the system interconnection is to connect them asynchronously. This will help maintain the system even if some Ac lines were out.

 

[advidana]

I think you should think again about anwering these types of questions. There is a lot of people out there that can use this information for bad intentions. Loose lips sinks ships.

 

[giuseppe]

Dear all,

interesting question dictated by the present emotions

One important question I always had, and that upset lot of people generally, is as follow:

"had the proper Testing/Pre-commissioning/Commissioning activities carried out?"

Testing: is to verify ON SITE that the single equipment/apparatuses has not being damaged by transport/storage/installation, and they are still perfectly functioning.

Pre-Commissioning: is to verify that every single System within the overall plan is properly working and reacting as expected.

Commissioning: Is to verify that the Overall Plan is properly working as designed.

In my quite long work experience I verified that often, due to budget restriction and Management skills inadequacy, the Project Management Team is reducing the attention right at this extremely important and delicate stage: Commissioning.

Most of the Management is convinced that Factory tests shall not be verified on site, that relays shall be only set to the designed values and that cables shall not be tested "because they're right, I can see it...!".

This mismanagement is often supported by the employment of under-qualified personnel (this is just for "yes-sir" purposes), adding problem to problem.

A distribution system that could not stand an overload (if it was it) or a localised short circuit, lightning strike, outage of one power plant, etc, is just a system born lame and destined to give troubles. Where the Experts were during the construction and Commissioning of this overall system? Or were they been "overhauled" by politicians in needs of a quickie...?

So far, I noticed a good job from everybody in this forum, even if sometime the truth is quite painful and obvious...

At the end, some operator will be the final responsible, the ones, of course, that could be sacrified without affecting the prestige of their bosses...Welcome to the real world of hypocrisy! But the problem will still there...

Cheers to all.

Giuseppe

 

[RobWard]

As a humble jobbing sparky much of what you are all discussing is a bit over my head, but it's interesting non-the-less.
When I was at college my lecturer recounted a tale of when he was "the man in charge" at a power station. He said that on one night another power station had shut down and demand kept rising. He and his colleague were watching the dial showing the output frequency, as the needle dipped towards the big red danger zone.
He said that they really really didn't want to have to press the "special red button" to isolate the power station. He explained that in doing so you relied on a lot of air breakers and emergency steam vents to operate or you ended up doing horrible damage to the turbines. At that point there was no-one in the country still able to fabricate a whole new shaft and they would have to get one from Japan at a frightening cost and the station would be out of action for a long time.
He explained that that put you under a lot of pressure NOT to press the button.
Fortunately as he and his colleague debated whether the needle of the dial was on the line or below it, demand started to fall and the frequency picked up.

Now, I suspect that things are a little different now (his tale must have been from 30 years ago) but you can imagine people being in a position where they really don't want to have to make the call....

"I love deadlines. I love the whooshing noise they make as they go past." Douglas Adams

 

[HiSet]

In my opinion,the effect of this outage could have been reduced or minimised had power system operators acted promptly to initiate manual load shedding or islanding schemes.Some where some body has goofed up for sure.

 

[scottf]

My 2 cents worth....

After all is said and done, I believe the investigation will show a chain of events and that there will not be 1 single cause of the outage.

From an industry point of view, I believe all the attention to the system will be a positive for the future. I believe utilities will be able to get rate hikes to pay for transmission upgrades/new lines. I also believe that this will spark a public debate over deregulation. In the end, I believe deregulation will be used as a tool to spark reinvestment in the transmission system (note that the rates have to accomodate this) and regional ISO's will be the tool to regulate reliability standards.

In the end, everyone's rates will go up slightly, but that's the price you have to pay for reliable power, in my humble opinion!

 

[ScottyUK]

I haven't seen it mentioned in the posts up to now, but what is the mix of generation types in the affected area? I know there are about six nuclear plants, but what are the rest?

Combustion turbines tend to react badly to dropping frequency because their compressor efficiency falls away so quickly. A drop in frequency of about 3.5% would cause the turbines at this site to dump load by disconnecting from the grid, as an example. If that takes place at a few sites, a cascade will occur unless the transmission system can shed load before the remaining generating plants dump load to protect themselves.

I don't know how a nuclear or hydro plant responds - I guess they are a lot less sensitive.

 

[peterb]

I had a rude awakening when I read the following in the IEEE PES PRSC report on Wide Area Protection and Emergency Control (talking about a WSCC disturbance on 10 August 1996)-
"The conditions leading to this incident built up over a period of 1 1/2 hours before the disturbance started. During this preliminary period three 500 kV lines in Washington and Oregon tripped out. Since these lines were not heavily loaded at the time, it was not recognized that the transmission system strength was being dangerously undermined with respect to its ability to withstand another contingency."

The discussion goes on, with a lot of other startling similarities to the present case. Seems that lessons weren't learned or the lesson wasn't adequately passed on.

 

[SidiropoulosM]

Peterb, good find. It will be interesting to compare the two events once the details of the recent one are better known.

 

[jstickley]

Based on what I've seen from preliminary disturbance reports (

http://www.nerc.com/pub_doc/PreliminaryDisturbanceReport.pdf)

and heard (from system operators), and from my knowledge of the Eastern Interconnect, my gut feeling is that there was likely a voltage collapse on the transmission grid, caused by large imports of power and triggered by the loss of critical generation and transmission facilities.

It will be interesting to find out the root cause of all of this, not to mention the fallout in industry. I hope that FERC, DOE, Congress, and others do not step in and make rash decisions and policies as a result of this.

That being said, I've gotta get on my for a second:

Government has given industry more than enough already, with open access and the development of power markets (as good as those things are) contributing as much to this as the physical outages that triggered the event (through the pre-event state of system operation).

I firmly believe that reliability policies are best left to those in industry who know their systems best. Contrary to what you've heard on the news, I believe that NERC is doing its job, and that utilities as a whole are doing everything they can (within the frameworks they are constrained) to prevent these types of things from occurring.

Anyway, that's my two cents on the issue. As I said, it will be interesting to find out root causes.

 

[gigahertz]

This type of thing happens because the system is at some point constructed for n-2 or n-1 conditions, and then consumption growth causes peak loading which exceeds the limits for n-1 operation (ie, less than n, no operation). The growth is allowed to occur faster than additional necessary infrastructure can be funded and built.

Regarding islanding, I think you will find that it is illegal for operators to disconnect from the grid. They must stay connected and ride it down through automatic shedding with everyone else. Without this, each operator could indepently decide to protect his own equipment as a priority and the system would collapse quicker and MUCH more often.

Basically, infrastructure growth must keep pace with load growth. If it doesn't, this type of thing will happen and we will get a lot of band-aids until it happens again.

 

[pennsy]

Here is my 2-cents worth. I heard today that the "System" was only running at 75 % of capacity when the outage occured.
I heard also today that our Electric Rate is going up( I live in Ohio, now) becuase they are NOW going to invest in new equipment.
Why havent they built a new power plant in my area in almost 20 years ? It sorta snuck out when someone mention drilling for oil in Alaska, and the response was OH we cant do that. Why NOT ?
Why cant we build new nuclear facilities with safeguards, that they all have now, The Perry Plant in Eastlake , East of Cleveland shut itself down no problem.
As far as I am concerned Wind power is just as the name implies, a bunch of hot air.

Thank You for allowing me MY 2-Cents worth.
Wm.Colt

 

[gordonl]

I'm in a large industrial on the fringes of the outage, but in it none the less. Internally my system islanded very quickly to protect our assets (a good 15 minutes before the lights went out), but as indicated by GHz this is a detriment to the overall system but I come first. (I'm using underfrequency loadshedding to seperate myself from the utility and control internal load on generation, and it works very well, though relatively frequently at least once a year) The small utility we're embeded in is a "slave to the master regulator (goverment agency), and although their system could easily island, they cannot take independant action. As a result we blacked out killing all local generation (except our own internal) and the only saving grace was a maint man in a remote station who managed to isolate one gen to station services, had it not been for that the system would not have had the ability to restart. (No black start generation in area)

What is better - island and keep some lights on or try and ride and fall hard. From my point of view power will come on a lot quicker from an island versus a desert.

A point of interst, I only ever hear my utility talk on N-1 contingencies, what is the practice out there?

 

[Skogsgurra]

This is getting more and more interesting.

I have spent many hours reading your postings and searching the net for additional information. And I am getting confused. But, behind the smoke, I can feel that it all is about finance having had a little too much to say and not having listened to the technical guys, those that have the necessary knowledge and the necessary insight. But that is what the world looks like today. Not at all unique with the electric. We probably have to live with this state of affairs.

My thinking is that we need very realistic models if we want to communicate better with the "accounting gnomes". Models that are updated in real time and that knows about every generator, every protection setting, every tap changer, every switch, every capacitor bank, everything. Having that model we can follow the development of problems and by testing "what if" also find out when critical situations are close.

By showing the result in easy-to-understand graphics I think that even the top management can understand when a dangerous situation is developing and take/make the right actions/decisions. It is far better to do it from insight than out of necessity, like now.

In the process industry, such a model is called an "observer". It runs parallel to the real process and much knowledge can be gained from such systems. A corresponding system for electric power transmission systems will of course be a huge system and needs a lot of real time data transmission, which has to be reliable. But still, I think that it could be done.

Do you know about any systems like this? Systems that cover large distribution networks and that are 100 percent correct and truly real time?

Would such a system be of any help?


Gunnar Englund