Relay & Control Panel Cost Reductions Project 16

[jlgordon]

Are there any other utility engineers responsible for finding ways to lower substation costs for 2003. I have been authorized/ordered to throw out our 100+ years of relay panel standards and start over with a clean sheet of paper on a new design for our relay, SCADA, and control panels to be built in the future. Not wanting to reinvent the wheel, I thought I would begin with switching from our custom everything to more off the shelf components. Has anyone blazed this trail already and if so, what are some of the problems to look out for? Our substations are used to serve both areas of large cities to rural farmlands in Texas using votages ranging from 12kV to 345kV. I have been authorized to utilitize microprocessor based relaying for both transmission and distribution class subs, versus the old Electro-mechanical relays. We have used one of the more common microprocessor brand relays for distribution class subs for the past eight years. Management wants even more savings. So starting in January my standards are gone and an empty white paper will be staring me in the face. All ideas are welcome, extremes included. I will even have a budget for prototyping and testing in a lab so "reliability" will be maintained. What are the most economical panel designs? We currently use 19" rack mount for distribution class relaying and custom punch steel for transmission class relaying. Is there anything else more economical? We are looked at having turnkey systems manufactured by switchgear manufactures to getting ride of our steel fixtures the breakers are previously installed with to pole mounted reclosures with a control box on the opposite side with all the relays.

 

[jlgordon]

fb599, what you're describing is exactly my worst fears. I am a big computer enthusiast from way back. I know just how rapidly advances take place, where microprocessor are involved. I also know what made some computer mnfs. successful and other not so successful. What I hope to be able to do is lay out the foundation that will allow the protection schemes to change while minimizing, down time, costs, reliability, ect. Today for example, when our utility upgrades an existing 64MVAR cap bank to 108MVAR, we wait until an outage is authorized and then we cut out a large hole in the relay panel where the Electro-mechanical relays were previously located. Then a custom-made cover plate is installed with cutouts for the new microprocessor relays. Finally, the new relays are installed and wired into the rest of the existing protection system. The whole process takes weeks.

This method is crude and time consuming. The construction personnel do not enjoy cutting holes in the panels, as there are many traps to fall into. The whole purpose of this thread is to look for a good way to allow for all the upcoming changes that are sure to come with advances in the power protection environment in the upcoming year, in as efficient manner as possible.

Over the last three decades, we have built several new substations using solid state or microprocessor relaying only to have catastrophic failures and / or a strong negative reaction from field personnel responsible for the operation and maintenance of our substations. Valid arguments were always made in each case and we have always reverted back to the old tried and true method we are all so familiar with. In several cases some very bright and innovative engineers have been forced to look for employment elsewhere. I think the key is to design a relay protection system that is functional, when initially installed, and has the ability to be changed as relays are discontinued. Preferably, without having to change out the whole protection system.

I think the way a PC is upgraded is a key component to the success of specific companies and I feel that it is also a vital concept that can be used to manage the costs of upgrading our protection schemes. How can we change a subset of a panel's protection scheme in the most efficient manner possible? I don't want to have to throw away a whole, custom made, steel panel with every change in the future. We had hoped that the rack mount panel would be the answer but those we have built, as of today, have not been acceptable for various, valid, reasons. We are now looking at using enclosures versus custom made steel panel or rack panels. Hopefully, we will find the modular replacement stratagy, I think we need, and eliminate the concerns / fears of the operations personnel about the rack mount panels used in the past.

Oh yea, I have to do all this for less money too!

 

[busbar]

At want point do you quit rerouting SIS wire and make all changes with flash-memory upgrades?

 

[peterb]

A couple of questions/points to fb599 & jlgordon's posts -

1. Where are you going to find the electromechanical relays to continue the "tried and true" methods into the 21st century?

2. Exactly how maintainable are these relays? Yes, you can actually touch and feel the hardware and you can see the parts move, but what is the cost of keeping this equipment in service? How much do you invest in training techs in this obsolete technology, and for how much longer?

3. In the mid 90's I completed a project that replaced a utility's electromechanical line protection schemes with microprocessor relays. In most cases, we were able to spec relays that fitted in the existing cutouts, with blanking plates taking care of the numerous empty spaces left over. Results were dramatic and are continuing - correct operations moved from 85% to 98%+ overall in the first year. Calibration drift dropped to zero and test times were considerably reduced. The fault location feature made instant admirers out of the operations personnel, whose line patrol efforts dropped considerably. Fault recording and oscillographic data provided invaluable information for fault analysis and correlation for system calculations.

4. Rack mounting is without question the most flexible method of accommodating relay equipment. I am familiar with large utilities that have used open racks for decades, with no adverse results. If your environment requires an enclosure, then these are readily available.

5. I, too, have done my time with electromechanical relays during the course of the last 30+ years, so I feel qualified to comment. The skill and ingenuity of the design engineers of the time are unquestioned, as evidenced by the wide variety and (mostly) reliable operation of the products that they produced. Having said that, there is no way that I would consider specifying an electromechanical scheme in today's world, with the exception of single-function devices such as high impedance differential relays.

 

[tomatge]

I'm interested in the comment made by fb599 concerning REFxxx relays, since we too have had problems with them (and I've heard of other utilities likewise). I gather they are prone to the software "locking up" and the relay simply stops working. However, they don't necessarily flag a watchdog to say so.

The Alstom C922 product has been withdrawn from the market - some problem with intellectual rights.

 

[jefffoley]

I would recommend SEL products as well. I have been commissioning systems (up to 500 kV) and the SEL product continues to become more prevalent. As another person mentined, once you learn 1 relay commands, you know all relays. I would stay away from SCADA systems and utilize communications processors and relays.

The new UCA2.0 "Goose" communications standard will allow all manufactuers of relays transmit analog and digital logic and metering data.

SEL offers turn key solutions. They build the panels, commission and provide a nice SCADA solution via wonderware.
Sel also can do the system study and most important the relay settings.

The biggest problem I have seen in the field is that "lazy engineers" do not take the time to understand the relay, do not apply the relay to the full potential and really screw up the settings. Having SEL do this will save you immensely.

SEL also has the best quality. In 2000, I commissioned 8 line (2 8 breaker ring bus+transformer+lines)

 

[HiSet]

Although REF xxx relay is very powerful yet it's too complicated and not user friendly at all.In fact It's SEF functionality is flawed and can cause undesired operation.I invite readers comment and feed back on REFXXX relays.

 

[rafish]

I have watched this issue develop for almost 20 years.

Utilities and government installations in my area would not consider using anything but electromechanical relays until SEL came along, and for good reason. While electronic relays always had the potential of being more reliable, none that could meet "utility-grade" reliability requirements were made.

Now virtually all of the government or utility projects I see use SEL. Many of your responders suggest SEI in 19" racks. That is my current practice.

 

[powerswitchgear]

No. The issue is standardizaton of technology. Who starts it , how long does it last and who runs it to make sure that it lives. Apparently all technology has life 3- to 50 years. Later this may be dead but we want our products to be live. So the cost savings is in self training(organization have staff that will work for generations). and there will be pay off.
The products die due to aging. Have them designed for not aging or aging very slowly. Are materials available to achieve this goal. Yes they do exist.Do we use all the time -No - why we are mortals and like products to be mortal. So let us say people like you can change for good -better and best. keep it up and you will be there.
Thanks

 

[miti]

What is the degree of protection used for the control & relay panels?

Are the control and relay panels situated in the same room?

 

[Bung]

I tend to agree that numeric relays in some form of rack mounting is the way ahead. But the major problem ahead is for the field staff - all those fancy relays, each with its own (unique) interface which will screw up[ your PC in a million subtle ways if you are not careful.

Like it or not, various forms of Wintel computers are pretty much the standard for driving the new relays, but virtually none of the programmers of the setting software have read and understood the Windows API documents. Or they are the result of a merger of several manufacturers, and are still engaged in internecine warfare to see which brand predominates.

To cap it off, most of the relays are way to complex. They do far too much for the average ues in distribution systems. At least Alstom affer some goood simple relays - pity about S1 and the compromises between the 20, 30 and 40 series (from UK (GEC?), Germany (AEG?) and France (Alstom?)). SEL are the classical example - who needs 6 different definite time elements for 3 or 4 functions (neg seq, e/f, o/c etc) in 6 setting groups on a distribution system as in the 351A?

So I think the if I were to start with a clean slate, I would be looking for simple relays with simple interfaces that do the job I want without having a whole heap of redundant stuff. Flash RAM (bit failures) and electrolytic capacitors seem to be the weak areas in the relays we have used. So be ready for them - have a good stock of spares on hand. You will need a good firmware tracking system to keep up with the ongoing firmware updates. Be prepared to freeze relays at a particular firmware level ,as changes to frimware mean yo have a whole new relay on your hands. Give your techs a brreak by limiting the nmber of different software packages they have to learn - stick to one or two manufacturers.

I could dribble on for hours, but I won't. TTFN


Bung
Life is non-linear...

 

[Milliamp]

Let's see... how many contacts do we need to clean on microprocessor-based relays? How many times do we need to do maintenance on mp-based relaying? Don't get in the trap of relying on any one manufacturer. For high-voltage protection we use two mfgs. relays.
Standardization of your panels and testing/installation procedures can help minimize costs and insure reliability.

 

[dede61]

I would suggest to take a look at

http://www.vamp.fi

I have gained great experience with these relays the last couple of years. They have excelent specifications and are very user friendly. The last project I done where two mobile substations (20" containers) 7 cubicles 24kV - 2500A complete with arc protection included. Price was lower than the most "BIG" suppliers where asking to preform less.

good luck in youre quest