Substation protection design 3

[jaydesai07ee738]

Hey,

I am trying to design a protective scheme for a 220kv/115kv substation.
The first thing I need to do is to select a relay manufacturer. I am confused between Areva, ABB, SEL and GE
I tried to find the pros and cons of each manufacturer but couldn't find much. Can you please help me here.
And ya most important thing, I am a grad student and fresher. So don't have any practical experience with relay. I have designed schemes on Midterms and Finals but have not worked on a project.

It will great if anyone with ample experience can help me.

Thanks,
Jay

 

[sonic02]

Hi,

Typically the requirements of the type of relays would be dependent on the Owner. They may have some standards preference for relays of a specific manufacturer. If it is older substation with no source of DC supply, it maybe using electromechanical relays. The protection design needs to take this into account and replacement may require the same type of relays.


Sonic02

 

[jaydesai07ee738]

No there is no such requirement. I do have a DC power and its completely my decision to choose a manufacturer. But the only thing is I should have a strong reason for using a particular manufacturer which I am trying to find without any luck. Every thing points to what the engineer's opinion about a specific company is.
I am asking because I am very fresher and so I am getting confused which one to use for my project.

Thanks.

 

[davidbeach]

Trying to start a religious war are you? ;-)

Familiarity, field support, what others in your area are doing, and like probably play into it at least as much as technical merit. A relay that is technically OK, with really good support is probably a better choice than a relay that is technically outstanding but has no particular support. Ultimately it's a choice you have to make, and keep in mind that you will be living with that choice for years and years to come.

We (the company I work for) selected SEL twenty years ago or so (long before I was there). We have invested considerable time and effort in developing a design and support infrastructure around those relays. As they've come out with new relays we've adopted those relays. To now change to a different manufacturer would be highly disruptive. We'd have to continue to support what we have for the foreseeable future and have to develop the same support infrastructure for the new manufacturer. Probably won't change until it is far too painful to continue with what we have.

The preceding paragraph could be equally true had a different manufacturer been selected. Others did make other decisions, some (generally the very large utilities) use more than one manufacturer, but even then they aren't reevaluating that decision very often.

People get used to the relays they use and see other brands as doing things oddly or more difficultly. I doubt that you will come up with an application so specialized that there will truly be a technical basis for selecting one over another. If it truly is a blank slate that you're dealing with then figure out which manufacturer will provide the best support in your location.

 

[jaydesai07ee738]

No sir, It's just a curious mind that's asking those questions.

Thank you so much davidbeach for a candid answer. That pretty much answers the question I had. Now will move on to the next part of the project. Selecting relays and getting to know the functions and how to use the outputs to generate the trip logic.

Thank you so much,
Jay

 

[FreddyNurk]

There may be an 'X and Y' factor for some of your protection for the scheme which might mean that you're selecting identical function relays from different companies just to avoid common mode failure as well.

In addition to that, there are other 'non-technical' considerations that may come into play as well. Certainly local technical support can be critical when the gear goes down and you need the manufacturer to assist, as davidbeach indicated, but there are other constraints, such as long term availability of product and the obligation of some utilities to head to the market and demonstrate probity in the selection of relays. Sometimes this takes the form of a supply contract for a number of years, sometimes its on a substation basis (makes it quite difficult if you've got a number of subs to manage but keep buying different relays).

What you're likely to find is that whilst you've carried out scheme design, the non technical issues take more consideration than you might have thought.

I'm not sure that anyone these days makes a 'bad' relay, which probably assists somewhat, but makes it harder in other aspects.

 

[davidbeach]

We've found (and we're far from alone) that the greatest risk of single mode failure is not in the relay hardware/firmware itself, but rather it is in the engineers that figure out the application and come up with the settings. It has been our experience that every thing we can do to make it easier to put the protection system together tends to reduce the likelihood of error. We certainly had relays fail, but never two at the same time. That's not two at the same time on the same position, but rather that is two at the same time system wide; same type or not. So we see no particular value in using different relays to increase reliability, but actually find that different relays, and particularly different brands, would significantly reduce reliability.

 

[FreddyNurk]

davidbeach, I can't say that I'm surprised. I only mentioned it as some utilities have standards on these sorts of things, and the OP hasn't mentioned if this is a private plant or utility. It wouldn't surprise me if other places also had standards that state the same thing.
You're right though, just as important is that all parties involved (design, installation, commissioning and so on) understand whats involved with the equipment. I've seen a number of different results in testing same equipment in different locations, which is somewhat concerning.

 

[ScottyUK]

"People get used to the relays they use and see other brands as doing things oddly or more difficultly."

Unless you buy Areva / Alstom relays, then you see other brands doing things more simply.

Joking aside, David has given a great answer.

 

[submonkey]

Hi Jay,

I would encourage you to select different brands for A and B
protections. It isn't just relay hardware failures you are
covering for when you use different brands. You are covering
for:

(a) Blind spots in relay algorithms.
(b) Misunderstanding of relay behaviour by the protection engineer.

Item (b) is important. An application engineer is unlikely to make
the identical mistake twice on two different relay brands. Don't think
that (b) can't happen. Relay firmware is incredibly complex and should
be assumed to contain bugs.

Make sure that the relays you select comply with appropriate standards.
It might be worth looking at the IEC60255 series. Pay attention to
electrical noise immunity, insulation withstand, etc.

Specify the relays with a healthy number of spare binary inputs and outputs.
The need for extra IO can occur late in the design stage.

Look at how the different vendors manage firmware. Make sure that
firmware revisions are available for download with clear and honest
descriptions of what changed, and why.

Download the relay manuals and see how much effort the vendor has put
into making the relay behaviour clear. As has been pointed out elsewhere
in the thread, engineer error is a likely cause of failure.

Have a very careful look at the relay configuration software. Make sure
that the software is robust, easy to use, can export files and settings
in a form which suits your business processes. In my humble opinion:

Enervista from GE is excellent.
AcSELerator from Schweitzer is excellent.
DIGSI from Siemens is clunky, but works pretty well.
PCM from ABB is absolutely awful (very big, very slow, very buggy, adds windows accounts, etc).

Don't take my word for it. Get copies of the software for each range
of relay and play with it. Think about how you will document your
setting work while you play with the software. Some software creates
logic diagrams which are hard to print, include in documents etc.

Good luck.

 

[submonkey]

Sorry - second (b) should say (a).

 

[Ceast]

3 thoughts

1) If your using 2 different manufacturers to supervise weak engineering, update your engineers.
2) We have never paid for a Schweitzer repair in 15yrs. They have warrantied everything so far...
3) Tech/engineering support for SEL relays is second to NONE!

Cheers
Chris

 

[davidbeach]

Ceast, I agree with your assessment of SEL, but I'm not sure the support is equally good world wide.

 

[DiscoP]

I also agree with David's assessment.

If you are not sure of the support capabilities, find out what other organisations in your region are using. It will be a good indicator that the local product support is there (vendors need local orders to be able to fund local support), and also you might also be able to talk with the other organisations about issues too.

We use about 6 different manufacturer's products in our organisation. If I asked peope in our group which relay is best, I will eventually get 6 different answers.

None are perfect. The support you can get when something goes wrong is the difference, and I think that would change in different parts of the world. This would be of even greater importance if you are new to this field.

 

[ptcabb]

Disclaimer, I am from ABB substation automation.

And to correct one error, PCM600 does not make any additional windows accounts.

You're designing a protection scheme. You're not determining what to buy, that comes way after. Your application is most likely not any different from what has been done before, so look for comparable examples and see what functions have been applied where in those situations. Then don't copy that blindly, but actually understand what functions are used and why. Maybe you need more. Maybe less.

Doesn't end at the protection side though. This is a networked world we live in. Functional integration is what most manufacturers want to sell you now. Do you want to add control functionality in your system? What means of communication are you planning to use? Does putting everything in one box (one extreme) fulfil your philosophy or do you still require electromechanical relays (other extreme)?

So the first thing to do is to define your application and YOUR requirements. Either by making new specifications or using existing ones drafted earlier. Then send your specifications to everyone, and let them offer you a solution. Why do that work yourself? Ask for type test certificates and make sure the functions you intend to use are tested according to the applicable standards. Or better yet, ask them to verify.

Just don't send 30 year old specs and expect a useful response.

Oh, and if you get attached to a single manufacturer, you'll lose a competitive edge. In the end, they're not so different in their tools and idiosyncrasy's. The world is to small for that and you should be able to work with all of them. Once your single solution supplier start suggesting you should put in your spec that you need flexible datasets, or multiple local / remote switches on the IED, or any number of vendor specific IED functions you should realize you're being manipulated into putting non essential things in your specification that lock out the competition.

 

[submonkey]

Hi ptcabb,

I apologise if my comment about PCM600 adding windows accounts
was incorrect. Perhaps I recalled incorrectly, and it is CAP505
which does that.

Does PCM600 add any services for running a background SQL server?

Thanks,
Alan

 

[ptcabb]

Hi submonkey,

That's quite ok, CAP505 did make the MicroSCADA account after all, that's true.

PCM600 installs MS SQL server 2008, which runs its own services and in addition a SQL Server (PCMSERVER) and server agent.

In the same line, IET600 (61850 tool) install its own SQL server (IETSERVER) and server agent instance as well.