What does SCADA do, and where is it located??

greenspun.com : LUSENET : Electric Utilities and Y2K : One Thread

I have a couplfe of questions that I have not been able to get an answer, I would appreciaate any response... I have read the SCADA sites but they seem to get into the technical and I'm too stupid and don't really have the time. If it has been answered please let me know the URL... Thanks

1. What actually does the SCADA "system" consist of and where is it located. 2. Is the SCADA in the actual "power" lines themselves, is it another line next to the "power" line that monitors the "power" line at certain stages, is it at transfomer stations, is it at distribution stations or is it at the generation sites??? 3. I understand that the SCADA is not necessary to generate and distribute power but is it the SCADA that actually monitors loading and unloading, and routes and re-routes the flow of power. 4. Would the loss of SCADA (due to loss of telecomm) diminish the ability of the utility to respond to loading and unloading of the grids in a timely manner? 5. If SCADA is out could this cause a cascading outage? 5. Is the SCADA what causes the safety net to trip stations off line for self-preservation? Thanks for any help..

-- Anonymous, March 03, 1999

Answers

I don't know any more than you do about SCADA -- but a Google search turned up this site, A SCADA Primer. I've only just begun to read through it, but it seems to cover the ground intelligibly. It's out of Australia, so it may take a little longer than usual to bring up.

-- Anonymous, March 03, 1999

In the not too distant past, all substation switching was done by a process where a dispatcher wrote a procedure known as a "switching order" and transmitted it via radio or telephone to a crew in the substation. The crew would sign off the steps and read the meters to the dispatcher to confirm that the steps were done properly and in the right sequence.

SCADA is an acronym for System Control and Data Acquisition. The advent of SCADA automated the switching and data gathering functions. At the substation, a RTU or remote terminal unit gathers process information such as volts, amps, megawatts, megavars, equipment status info (open/close, on/off etc.) This is done using either electro-mechanical transducers with an analog milliamp/millivolt output (dumb devices) OR in some cases using a communications protocol (IEDs or Intelligent End Devices). The RTU concentrates the data and transmits it to the dispatch office.

At the dispatch office, a SCADA Master Unit collects the data from multiple RTU's and displays the data on the screen of an operators terminal. The dispatchers/operators can see status, alarms, metering etc. They can also initiate control commands to open/close, ON/OFF, devices in the substation. The commands are routed by the SCADA master to the RTU in the proper substation, where control relays operate as switches to perform the desired task.

Sometimes, the data of a SCADA master is used to calculate and schedule automated switching for applications like wide area voltage/var control. These can be programmed by the end user, and are a potential source of Y2K errors. My utility is replacing the SCADA master hardware and code, and the RTU's have been tested OK. No application programs our RTU's use Date.

Hope this meets the test and straddles the fence between too vague and too detailed.

-- Anonymous, March 03, 1999


I forgot to mention. The SCADA system is not required to operate the system and deliver power. It wasn't that long ago we didn't have SCADA. The April NERC drill is to test and practice the communications necessary to confirm that the operators/dispatchers can operate blind and lame in the event of a telco or other failure that could disable SCADA. Men in substations, reading meters and operating manual switches will (on 4/9) once again be the eyes and arms of the dispatchers.

-- Anonymous, March 03, 1999

That's a pretty good consise description of SCADA.

Sometimes it is described as "Supervisory Control And Data Aquisition".

It might be worth noting that SCADA systems are unique to the Power industry. They are also used in many manufacturing, automation and telecommunications applications.

--aj

-- Anonymous, March 03, 1999


AJ -

I think you meant to say that SCADA is NOT unique to the electric industry. ;-)

-- Anonymous, March 03, 1999



It might be worth noting that SCADA systems are NOT unique to the Power industry. They are also used in many manufacturing, automation and telecommunications applications.

--aj

(Thanks Rick)

-- Anonymous, March 03, 1999


AJ,

You caught my freudian slip. Supervisory not System. Maybe I am harboring a subconscience adversion to pointy-haired "Supervisors"?

-- Anonymous, March 03, 1999


Thanks for all your responses... It has cleared up a lot. I do have a couple of more questions.. I haven't been able to get to the primer yet..

1.Is SCADA hooked up in some utilities so that there is no human intervention? If it gives bad readings does it shut down the system automatically? Or is there human intervention required?

It seemed to me that the instances that I read about (San Francisco and the NorthWest in 1996), seemed to trip/shut down much more rapidly than would be possible with human intervention?

2.Could a loss of telecomm or a Y2K bug in the SCADA skew that data on the displays and cause a controller to make the "wrong" decision based on this bad info and cause plants to trip-safe or shut them down for safety? It would seem that it would make monitoring the system a fun time, if problems do occur.

-- Anonymous, March 03, 1999


Mark,

1.Is SCADA hooked up in some utilities so that there is no human intervention? If it gives bad readings does it shut down the system automatically? Or is there human intervention required?

I preface this by acknowledging my knowledge is limited to my utility, but I can envision no possible exceptions. Others can comment on whether this applies universally.

1. Yes and No. The concept of distributed intelligence is applied in protection and control of utility system. Certain applications where global knowledge is necessary may actuate by SCADA without human intervention. These cases would involve wide are voltage/var control (closing cap banks to raise system voltage, tripping them to lower system voltage) and would not involve interrupting power (merely conditioning it). The decision process is in terms of seconds and minutes.

SCADA actions that would disrupt you power, eg. tripping circuit breakers is done with human intervention for switching. Devices that protect the power system from short circuits (faults) automatically open circuit breakers by monitoring current/voltage flows and are located as close to the device (breaker) as possible. They must operate quickly, in the order of 1/4 in some applications. There is no time to incorporate wide area SCADA communications in such decisions. The intelligence for this high speed tripping is distributed, not centralized (located in each substation with current/voltage inputs for each breaker). The devices that initiated the SF outage are these protective relays which I have mentioned testing with no major and few minor problems (or Y2K READY).

The issue of "bad readings" is essentially the same as "no readings" and will be considered in #2 (below).

2.Could a loss of telecomm or a Y2K bug in the SCADA skew that data on the displays and cause a controller to make the "wrong" decision based on this bad info and cause plants to trip-safe or shut them down for safety? It would seem that it would make monitoring the system a fun time, if problems do occur.

Yes, bad data could cause an operator to make an improper switching decision. Would/could a Y2K bug cause this? Remotely possible, but highly doubtful. I think that a problem would more likely cause a reading to go entirely blank. I say this with confidence because the SCADA system in our company is "forever" having individual data points (and sometimes groups of data points) go blank. This happens when the RTU (point of data gathering) goes "out of scan" and is temporarily not communicating with the master station at the dispatch center. The Y2K error would have to "trick" the complex communications protocol between the RTU and SCADA master perfectly or be rejected by error checking (again causing a blank reading).

This would not cause an automatic trip, because as I mentioned in #1, that function is not a SCADA responsiblity, but is handled locally at each substation by the protective relays. The knowledge that the relays provide a high (supreme) degree of protection and that they operate independently of the SCADA system means that no operator EVER uses human intervention to trip a device that is out of scan. (Even if they wanted to, most times the failure of the remote data gathering is coincident with a loss of the ability to do remote controls - they would'nt have the means to trip from the dispatch center).

Yes, this creates a "fun" situation for operations. They call the SCADA technicians to fix the problem so that they are no longer "blind". This is of high but not great urgency because of the protective relays. SCADA is not necessary for the power to keep flowing. The dispatch center will not know how much, and might not know if a short circuit caused a relay to trip a line - but that is not a problem since customers are very cooperative about calling when the lights go out (grin).

If this were to occur widespread and simultaneously, this would cause great concern (but no outage). Even though we have no test results indicating a need for it, contingency plans call for manning of substations to provide the "eyes and arms" in the event the SCADA system fails widely. Remote communications to provide the manual data acquisition and control (switching) are being tested and practiced in the April drill.

So, in my mind the protective relays are ORDERS OF MAGNITUDE more important than the SCADA system. The relays are complex devices that only in recent years incorporated microprocessors. They have some date/time/sequence of event functions, but the protection and control aspects do not use date and are not impacted by Y2K rollovers.

-- Anonymous, March 03, 1999


NOTE TO SELF, PROOFREAD BEFORE SUBMIT!!!

In the answer to 1. of previous post - 1/4 should have read: 1/4 cycle, or approx. 0.00415 seconds - 4 milliseconds or less.

Sorry.

-- Anonymous, March 03, 1999



Cl,

-- Anonymous, March 04, 1999

Cl,

-- Anonymous, March 04, 1999

Cl,

-- Anonymous, March 04, 1999

Cl, Thanks for the information, I appreciate it..

-- Anonymous, March 04, 1999

Moderation questions? read the FAQ