Any new technology out there for power monitoring? 3

[bdn2004]

We would like to monitor volts, amps, power, etc at many other places on the power system at this plant. Power monitors are of reasonable costs and do a lot of cool things...but the installation of CT's and PT's would greatly drive up the cost of installation.

We recently installed a new dc metering system that used an ingenious non-obtrusive fiber optic sensor to monitor the dc current. Is there anything available like that for use on an AC system? Does anyone have any experience with such a thing?

 

[Zogzog]

You may be able to use your existing trip units on your breakers, what type of breakers and trip units do you have?

 

[jraef]

What kinds of currents are you wanting to measure? 10A, 100A, 1000A, 10,000A? 3 phase, single phase?

Keep in mind, if you are wanting to measure plant-wide power, you don't need PTs everywhere, only one place. Unless you want to adjust your readings for voltage drop, in which case you still only need them in select areas.

Current measurement still needs CTs. There are "better" ways of doing it (Rogowski coils for example), but not cheaper. Then you have the issue of acquiring the data, which usually means transducers and some sort of communication system. Several manufacturers are addressing this issue with combination devices. I used to work for Siemens, they have a device called a Simocode that is such a combo unit. It has a built-in 3 phase CT, with or without a PT option, a processor to do integration (V/A/W/PF/kWH etc.) and a comm processor to put it up to a Profibus network. There are several other products like it on the market.


"If I had eight hours to chop down a tree, I'd spend six sharpening my axe." -- Abraham Lincoln
For the best use of Eng-Tips, please click here -> faq731-376

 

[bdn2004]

I'm working on a SCADA functional specification first draft that will get the conversation started. This specification will be handed off to real experts, but I've got to get the basics down. The plant's main concern is to find out what happened quickly, so as to get back on line as fast as possible. We are going to monitor many different potential tell tale spots on literally hundreds of transformers, hundreds of pressure and temperature switches, overvoltage, overcurrent, switch positions, etc.

I'm no expert obviously, but it seems to me that when putting the puzzle together of what happened that a history of what was going on with the voltage and current at specific points would be beneficial to know.

This is a metals plant that uses many large phase shifting (as many as 9) and rectifying transformers that operate in parallel to provide the AC source for the rectifiers. An alarm in any of them will shut down the whole line.

For example say a temperature switch on one of the transformers trips, and then trips the main breaker, wouldn't the voltage and current at that point be beneficial to know?

And I don't know what the trip units on the breakers are...the breakers themselves are large oil circuit breakers in a switchyard. In general they have old style dial type IAC relays protecting them. What specific info do I need?

 

[rbulsara]

We would like monitor volts, amps, power, etc at many other places on the power system at this plant. Power monitors are of reasonable costs and do a lot of cool things...but the installation of CT's and PT's would greatly drive up the cost of installation.

is an oxymoronic statement as that is what is needed for what you want to do, the right way. Either you have a need to know something and willing to pay for it or not.

What Zogzog indicated is a very good option for new or retrofit installation but of not much help with existing systems. Even there most mfrs have a product that works reasonably well with their own products but not so with other mfr's devices, without some sort of interface or drivers. Many are not good with making their software backward compatible, as new product lines come out.

This obviously is not a project for free internet advice. Get a group of professionals, if you are not already part of one, together on board and work with them. You need to breakdown the task in smaller chunks or multiple layers working top down. Starting from service, feeders, branches and devices and so on.

SCADA is not only expensive to install, it is also costs to maintain it keep it up-to date.

Rafiq Bulsara

http://www.srengineersct.com

 

[racobb]

Check to see if the oil circuit breakers have an extra set of CT's already on them. You probably already know that DNP is a good comm protocol that many vendors support.

Alan

 

[bdn2004]

rbulsara,

I realize the limitations of this free forum.

I also have found by experience that coming in with a skeleton of a path forward - even if it is wrong - in the form of drawings, Powerpoints, suggestions, questions, preliminary estimates...is way better than going into a meeting with a lot of busy, important people and saying... "So what do you want to do? Where do we start?"

I've found people like to mark up things, and they'll say no don't do this but do this. Eliminate this. Put this in But they have to have something to mark up.

That is my only objective here.

I'm amazed the quality of the answers of almost everything I've ever posted here, including yours of course. I'm just trying to do as much homework as possible, before we go down a path that will end up costing way to much and never get installed. And to be truthful that path on this project has already been tread...I don't plan to repeat it.

 

[rcw retired EE]

Based on my experience running maintenance and operations in a metals plant, here is what I'm guessing you are looking for:

1.Alarms and SER- A way of quickly telling what happened so you can get back on line. The operators might be able to save time if they knew what tripped the breaker, without having to run out to various relay panels.
2. Analog Data - A means of gathering data to do a post mortem to analyze what went wrong..
3. Proactive System- A system that could alert operators to correct a problem before the trip occurs by monitoring critical readings (temperatures, voltages, amp loading, oil levels, cooling air filter DP's, auto-tap changer operating sequence timing, etc).

Those criteria are roughly in order by cost.

1.- Alarm system that is a little more sophisticated than an alarm panel and trip targets on relays. Look at adding trip indicating relays to each trip circuit that close a contact when trip current flows. Wire the outputs, along with all breaker and lockout relay aux contacts into a Sequence of Events Recorder or a PLC/DCS system to record the trips. A PLC or DCS could have "expert" system programming to use the inputs to determine the fault location, probable cause, recommended recovery action, etc.

The transformer and rectifier low level, high pressure, high temperature alarms could all be wired into this system,

2. IMHO, voltage and current records will not be that useful in a metals plant for troubleshooting. Nice to have, but not critical because most of the problems will be hard failures of cables, transformers, rectifiers, or other more mundane trips. It's not like a transmission line or a generating station with sophisticated relaying schemes that require lots of data to analyze when a false trip occurs. Monitoring voltages and currents on a few main breakers may be all that is needed. (Unless harmonic problems start occurring.).

3. This level adds analog I/O to read and record transformer and rectifier temperatures, amp loading, tap changer operation or other variables your experience indicates is critical to your plant.

Define what you want to do, then start looking at devices or systems to do it. Start with a one line, and hit the big equipment. Think of what information you would like if you only had 30 minutes to get back on line after a trip. Start a list of data points based on the one line.

Decide if it will be data only, or do you want to do remote control also.

Decide if you will be using remote I/O with modules in each breaker/ transformer cabinet scattered through out the plant, or centralized I/O with field wiring back to the central location. Are there existing fiber networks that could be used?

Just some ideas to get you going.

 

[bdn2004]

rcwilson,

Exactly....it's as if you work here! And good advice about power monitoring, its probably why they don't get that data now.

I'm doing what you are suggesting: developing a single line, an I/O list, and a simple sketch of where various equipment we are monitoring is physically located. And as you say, hitting the big points first. The vendors I've talked to have told me they will lay out a proposed system architecture after they can get that info and give me a budget quote on the electronics needed.

This project has been proposed before and went down in flames because it was way too massive,way too expensive and in my opinion way too complicated. These kinds of projects are difficult to prove their value to those holding the cash - especially when ultimately what you are trying to protect against, is not happening very often and when it does the problem gets solved eventually without much damage despite the anxiety it caused.

BTW, you mention cable failure is a biggie...and they have had cable failures in the past. What can you do to determine this is indeed what failed?

 

[rcw retired EE]

You probably can't automatically deduce a cable failure unless the system has differential relaying on the cable circuit. (A refinery I worked in used pilot wire differential on all critical feeders and ties. We could quickly deduce cable failures.) Usually, the system will just get a signal that a ground relay tripped a breaker. You can't tell differentiate between a fault in the cable or in the load unless downstream relaying can provide further discrimination.

But you don't need to know right away if it is in the cable or the transformer, You just need to know what zone has failed, how to isolate it and how to safely return power to the unaffected equipment. The feeder cables will be in the same zone as the transformer it is feeding. That zone can be isolated by opening the secondary main breaker on the transformer and the primary breaker.

If the cable feeder daisy chains out to several transformers, install fault detectors on the outgoing cables at each junction and wire the output contacts to the SCADA. You will be able to tell which cable section has a problem.

Odds of getting approval to go ahead with this are not good unless it is part of a larger project or a protective relaying upgrade. The payout is not there on an existing plant, until a major failure occurs and the operators have problems recovering. I tired, but couldn't sell the new system. Management opted for a minor alarm system upgrade, a better training program and detailed operating instructions with troubleshooting flow charts and check lists. The same logic we put together for the “Expert System” was put in the flow chart for the operator to follow.

Selling a SCADA system because it helps you get back on line faster only pays out if your metal potlines are going to freeze in an hour. But if you can sell it as a protection system upgrade if the data will help you extend maintenance intervals or squeeze more life out of older equipment.

 

[jimioy]

We use Schneider Electric (Sq D) Power Logic metering on our 25+ facilities. We do the installs, use Ethernet or fiber and set it up on our network. The meters are also getting inputs from the water and gas meters as well.

Jimi Young - civilian
184th CES/Electrical Shop
Ks Air Nat'l Guard
McConnell AFB, KS