GENERATOR DIFFERENTIAL CURRENT PROTECTION TRIP ISSUE 1

[TONY2023]

Hi All, I need some help. I have a genset that has a differential protection scheme as per attached one line. We have a set of CTs on the bus side with Ansi or American style current transformers, and a second set of CTs that are IEC or European Bar style CT. In an ideal situation, two sets of CTs shall have CT polarity facing away from generator for the differential current protection scheme to work properly. Since the IEC bar style doesn't have CT polarity, the relay still displays generator neutral current value correctly. However, when I ran the genset with load, the relay tripped on differential current tripped. I saw the relay displayed differential current was about two times of restrained current or phase current.

How can I wire the two sets of CTs so that genset wouldn't trip on differential current? Since one set of CTs has polarity and the second set of CTs doesn't have CTs polarity, how can we make them work on the differential scheme?

Thank you for your help!
Tony

 

[davidbeach]

All CTs have a polarity, it may just not be marked. But the CT polarity is merely a detail to get things connected correctly. What matters is which end of the CT winding is connected to what terminal of the relay.

You need to indicated which end of each CT is grounded and which end is connected to the polarity terminal of the relay. There are no connections to the center of a CT. Once you have that drawn out to match the actual installation the source of the problem might be blindingly obvious.

When one this sentence into the German to translate wanted, would one the fact exploit, that the word order and the punctuation already with the German conventions agree.

-- Douglas Hofstadter, Jan 1982

 

[edison123]

Just reverse one set of CT's.

Muthu

http://www.edison.co.in

 

[davidbeach]

Ah, but which one? The generator relays I'm most familiar with require that one set of CTs, normally installed at the neutral end, be grounded toward the system and the other set, normally installed at the terminals, be grounded toward the generator neutral. In that case, measuring the current into the generator. But in split diff hydro machines, that first set of CTs have to be installed at the terminals, where the entire current is available, and the other set, with a fake CTR setting in the relay, installed at the neutral end; in this case the diff is current out of the generator. Many ways of getting it to not trip on load and trip on internal faults, but far fewer ways to also have it get all of the other generation protections correct at the same time.

When one this sentence into the German to translate wanted, would one the fact exploit, that the word order and the punctuation already with the German conventions agree.

-- Douglas Hofstadter, Jan 1982

 

[che12345]

"..I havmillie a genset that has a differential protection scheme as per attached one line. We have a set of CTs on the bus side with Ansi or American style current transformers, and a second set of CTs that are IEC or European Bar style CT.."
I have the following opinion for your consideration.
1. ALL Cts irrespective of IEC or ANSI design have polarity, i.e. the relationship between the primary and the secondary, e.g. P1 P2 , S1 S2. When current is flowing IN from P1 to P2, the secondary current is flowing OUT from S1 to S2.
Note: This can be easily determined with a small battery and an analog centre zero dc mA meter.
2. Determine the polarity of all SIX Cts.
3. Note: It would be more difficult to determine whether IEC and ANSI are compatible. Attention: usually they are NOT compatible. However, most relays has some provisions to circumvent this "inherent" problem.
4. Wire the circuit according to the relay manual. Observe the polarity strictly, usually shown with dot/square/ + notation. Attention: Observe that there is ONE and only ONE earthing point for all six Cts.
Che Kuan Yau (Singapore).

 

[davidbeach]

It's not true that all six CTs have to be grounded ("earthed") at one point. Many relays leave 3-phase sets of CT inputs isolated from each other (usually ever CT input is isolated) and it's certainly permissible to ground each set of three CTs separately. Back in the EM relay days that wasn't the case, and yes, then, just one ground for the whole lot. But as long as there's no circuit that can involve CT secondary current flowing on ground it's allowable to ground different groups separately from each other.

When one this sentence into the German to translate wanted, would one the fact exploit, that the word order and the punctuation already with the German conventions agree.

-- Douglas Hofstadter, Jan 1982

 

[che12345]

"...Back in the EM relay days that wasn't the case, and yes, then, just one ground for the whole lot. But as long as there's no circuit that can involve CT secondary current flowing on ground it's allowable to ground different groups separately from each other"

1. Irrespective of EM or micro-processor relay, ONE point earthing is a MUST. Multiple earthing is a sure recipe for problem which may not be evident at the moment. It does not occurs during normal heathy operation. There are numerous reports that this happens on heavy through faults, when the problem surfaces !.
BTW: In this case, they are IEC and ANSI Cts, which could be even more complicated.
2. Attention: Observe ONE point earthing strictly, it costs nothing.
Che Kuan Yau (Singapore)

 

[sushilksk]

From this, CT polarity is the probable cause of trip. Checking CT polarity and correcting is the solution.

Further, regarding single point earthing for CT secondary circuit: Generally, one side set of three phase CTs make has siingle point earthing. And other side set of three phase CTs has its own single point earthing. Practically, multipoint earthing cuases trip when there is ground fault in system. In normal balanced system it does not cuse much spill current.

Regarding IEC/ANSI CTs: There should not be differential current for normal currents below CT continuous rating, it may be a issue when there is through fault.

 

[catserveng]

What kind of protective relay do you have installed?

If your fault recording is properly setup then you should have an oscillograph or data from the fault. Using that info will help you identify the actual reason for the trip and point you in a direction to correct the issue.

I have worked on a lot of genset systems where the switchgear had ANSI donut style CT's and the generator neutral side CT's were IEC style bar type, with both sets having a 5 amp secondary likely they will work if properly installed. You may have to consult an experienced protection engineer to make sure the differential settings will allow for differences in the secondary outputs of the two types of CT's, but most modern relays have the ability to compensate for those differences if properly setup.

Does the unit trip on differential right after the breaker closes, or as load comes up at some point?

More details will get you better answers, hope that helps.

MikeL.

 

[cranky108]

CAT, that is a point that I have not seen above. All CT's must have the same ratio, unless the relay takes that into account. Thus all CT's are 5 amp CT's, or all 1 amp CT's.
A single point ground on each set is required.
As far as looking at the relay recording, most relays that I have dealt with need a potential reference. If the relay does not have that the recording becomes somewhat jumbled, or unreadable.

Lesson, even if you don't use any potential functions in a relay, still bring a potential to it as a reference.

 

[catserveng]

Cranky,

I was going by his drawing, shows two sets of PT's.

This is an exercise I have had to go thru many times over the last several years, made me REALLY appreciate modern multifunction protective relays. Had multiple customers running gensets, mostly landfill or digester gas plants, that many times after a major overhaul that included a generator removal and reinstall, trips on differential after return to service were quite common, mostly due to wiring errors on reinstallation, sometimes due to damaged CT's, and a couple times real trips due to actual generator tail end problems.

A phasor diagram showing the problem(s) helped many times and made troubleshooting and correction a lot quicker. Was also very helpful in documenting the issue, as many times by the time I got involved the unit had been down for more time than expected and folks were upset the unit wasn't back online.

I serviced several older sites that still had Basler BE1-87 relays, so I setup on old Basler GPS100 relay with differential as a test tool, helped cut troubleshooting and downtime a lot. Also used it as an example of the benefits of a newer relay to sell protection retrofit jobs.

MikeL

 

[TONY2023]

Thank you for your feedback. After further review and discussion with generator application engineer, I found out that they grounded all the CTs secondary winding S1. So, I was able to convince him to allow me to change neutral CTs wiring by using three S1 wires connected to relay SR889 per this attachment, and grounded all secondary winding S2 wires.

Based on this attachment wiring, do you think is the polarity correct and I will not have any trip on differential?

My next question is the set of CTs on bus side. Since polarity H1 is facing away from generator, will current flow into H2 and out to X2 to go to relay? Or current flows from H1 and out to X1 to go to relay? The reason I ask because I see current displayed correctly value on relay per wiring on hand sketch. If I switched wiring from X2 to 1 on shorting block STB10, and X1 to 2 of STB10. The same for CT5 and CT6 wiring. I got a weird current phasor and values close to 1A.

 

[davidbeach]

Your neutral end CTs are the problem. For each one of them, swap their connections to X5. They need to be pointed toward the generator instead of a way from it.

When one this sentence into the German to translate wanted, would one the fact exploit, that the word order and the punctuation already with the German conventions agree.

-- Douglas Hofstadter, Jan 1982

 

[edison123]

Swap 12&9, 11&8, 10&7 and see what happens.

Muthu

http://www.edison.co.in