Transformer-Arrester Connection on Pole-Mount Transformers 4

[joozu6]

I work for a electrical distribution utility. This past season we have had over a dozen transformers that were struck by lightning, resulting in the transformer spliting from the top of the transforer close to the primary bushing and down to the tap changer (we use dual voltage 7.2kV/14.4kV transformers). Our current method of connection is from the overhead line, through the cutout and then to the primary bushing. The arrester is connected by connecting a jumper from the bushing wire to the top of the arrester. The arrester lead is grounded to the tank ground.

It seems to me that this connection will incourage a ligtning surge to go through the bushing, resulting in our tank-splitting problem. Does this seem logical?

 

[RalphChristie]

The arrester have to be as near as possible to the equipment to be protected.
The arrester have to be between the incoming line and the equipment to be protected.


Lmax = {{v [ BIL / (1 + Mp)] – Ures } / 2 x S} - (H arrester + H structure)


Where:
Lmax : max allowable distance between arrester and equipment(m)
H arrester : Height of arrester (m)
H structure : Height of structure(m)
V : velocity of wave propagation (m/ms) » 300 m/ms for overhead lines
150 m/ms for cables
BIL: Basic insulation level of equipment to be protected (kV)
Mp : protective margin required : Minimum IEC recommendation = 20%
Ures : Residual voltage of arrester (kV)
S : Steepness of incoming wave (kV/ms) » 2000kV/ms

Protective distances
kV max distance
11 on trsf tank
22 on trsf tank
33 3m
44 5m
66 6m
88 6m
132 7m
275 18m
400 24m



Regards

Ralph

 

[busbar]

Basic overvoltage-protection concepts are partially addressed in IEEE Std C62.22.1-1996 …Connection of Surge Arresters to Protect Insulated, Shielded Electric Power Cable Systems

A significant improvement can be achieved by reducing the “open air” cross-sectional area enclosed by the arrester, jumpers and transformer bushings/can.

 

[tinfoil]

Our system is 12470V (7200V L-G).

Our practice is to mount the arrestor beside the transformer's 'cutout' disconnect switch, with the 'hot' lead connected in parallel with the source side of the cutout, and then the arrestor is connected to ground below the transformer's ground connection, and then sharing a common ground wire down the pole to earth.

The cutout is usually ~1m from the transformer bushing.

It was felt that this maximizes the arrestor's ability to 'shunt' the lightning surge around the transformer.

 

[tinfoil]

You may also want ot review your grounding practices. An arrestor's ability to redirect energy away from other equipment is only as good as the overall impedance to 'true earth'.

If the local connection to earth is poor, the near-zero impedance of a MOV arrestor will still not be a very good shunt, since it is in series with the ground wire and earth grounding.

We have found that simply using a single ground rod is insufficient if the local ground resistivity is high (soil conditions are rocky, etc.).

 

[joozu6]

Tinfoil, that seems like a logical connection.

One of the reasons I was told we connect from the cutout to the arrester (in series) is that the cutout will blow and indicate that the arrester needs to be changed out. Otherwise, the arrester may be bad and no one ever knows it. Sometimes the arresters that have blown look normal from the outside (from what I was told).

With respect to the arrester ground lead connection, we currently connect it to the tank ground. I also thought it would make more since to connect it directly to ground, but I read some previous posts that said connecting to the tank would reduce the induced L di/dt voltage between the windings and the tank inside the transfomrer.

Any thoughts?

Thanks for all of your responses.

 

[stevenal]

I'll back Ralph's advice here. Tinfoil's 1 M lead length (2 M from the bushing?) on the line side sounds very long. And the insulation to be protected is that between the bushing and the transformer can. True earth is a long way away (high impedance to the high frequency components of a wavefront). Connect between bushing and can keeping leads as short as possible. Might want to look into secondary protection as well.

 

[joozu6]

Isn't the idea to have the arrester drain the surge before it hits the bushing? Seems like you would be accomplishing that by Tinfoil's design. The impedance should be lower through the pole ground than through transformer.

 

[rbulsara]

trying undestand original post...

What is meant by transfomer splitting? is the case splitting? or the winding is blwon to pieces or both?

Sounds like the arrester never worked, all energy is passing through the transfomer case or winding.

tinfoil's point equally valid. You better have good earth connection to dissipate the energy. Current has to compelete the circuit and it will take path of least impedance.

Although you want to connect to the 'can' but not depend on the 'can' to carry the surge current, rather have a solid condcutor all the way down to good earth electrode system. (using a pad or lug on the can as junction).

 

[busbar]

USDA-RUS bulletin 1724e300

4.3.5 Surge Arresters Transformers, regulators, and other substation equipment are particularly sensitive to transient overvoltages. For the highest degree of equipment protection, surge arresters should be installed as close as practical to the equipment being protected. In most instances, power transformers can be furnished with surge arrester mounting brackets to facilitate installation. Separate arrester stands can also be used, or the arresters can be installed on adjacent switching structures. For voltage regulator applications, the surge arresters are normally installed directly on the regulator tanks.

When power transformers are protected by fuses, it is recommended that transformer surge arresters be connected on the line side of the fuses, as close as practical to the power transformers. This will minimize the stress on the fuse and help avoid partial melting of the fuse link when the surge arrester responds to a transient overvoltage.{/i}

 

[joozu6]

Transformer splitting meaning the case split open. We've had over 12 react the same way this year. I'd say it's time to change our connection.

Thanks for your help.

 

[stevenal]

joozu,

No the idea is to prevent over stressing the transformer insulation. Adding lead length (inductance) to either the line or ground side will increase the voltage at the can. See Ralph's post.

 

[rbulsara]

Insulation is not failing, the transformer can is coming apart at seams...sounds like stress or uneven expansion caused by overcrrent/heating..

Change the connectins such that current does not pass through the can's skin, but through the conductor. Although attach the conductor to the can so can is grounded too.

 

[jghrist]

I agree with Ralph and Stevenal. The shorter the leads the better because surges travelling through the leads impose an additional voltage on the transformer because of L·di/dt.

The transformer "splitting" may be caused by high fault currents on a failed transformer rather than by the surge voltage directly. Backup current limiting may be used to prevent this. See

http://www.sandc.com/webzine/021003_1.asp

http://www.sandc.com/edocs_pdfs/EDOC_001654.pdf

http://www.cooperpower.com/Library/TheLine/pdf/99035.pdf

 

[RRaghunath]

I think most important point is connecting the surge arrestor directly to ground through shortest and adequately sized leads. The ground means true ground and as others suggested it may be necessary to go in for more than one driven earth electrode to obtain low earth resistance.
Equally important is the rating of surge arrestor. If we consider the 11kV system to be solidly grounded type, a 9kV rated surge arrestor should be appropriate and the type shall preferrably be metal-oxide, gap-less.

 

[alehman]

Just curious, are these dual bushing transformers? If so, where is the grounded primary lead connected?

I wouldn't be as concerned for the actual impedance to earth as the total impedance between the arrestor and the transformer. You are trying to minimize the transient voltage seen by the transformer, therefore the arrestor ground should be connected close to the transformer tank ground.

The arrestor voltage rating is important. Also make sure your fuses and arrestors have adequate interrupting ratings.

 

[jghrist]

While it is helpful to have a low resistance earth electrode, the important factor for transformer protection is limiting the surge voltage across the transformer insulation. Minimizing the length of the lead from the top of the arrester to the bushing and the bottom of the arrester to the transformer tank (at the bracket where the arrester is mounted on the tank) does this by minimizing L·di/dt in the leads. The voltage across the transformer insulation is the voltage across the arrester (Front-of-Wave or discharge) plus the L·di/dt drop in the leads. The voltage to true ground of the earth electrode is imposed on both the arrester ground and the transformer tank, so it does not increase the voltage across the transformer insulation. Low earth electrode resistance will reduce surge voltage on the line and improve overall line performance.

 

[2dye4]

Also I would think the connection between the
tank and the local earth ground must be very good.
Otherwise there is a high potential between the primary
and secondary which is surely well grounded at the load.
So inadequate pole ground could contribute

 

[jghrist]

The primary ground is connected to the secondary ground. If the secondary is well grounded, so is the primary and the arrester.

 

[cuky2000]

See the enclose picture used by many utilities similar to the case described.

Notice that there is not shield wire to protect from direct or near lightning strike. Traveling wave could create severe overvoltage at the transformer primary side resulting in component damage.
Possible mitigation strategy is considering installing line arrester and/or OH shield wire.
The table below shows the line performance against the number of flashover for different configurations

Summary of results of the studies at
100 Ohms footing resistance

Case Configuration Total Flashovers Per 100 km/yr
I No Protection 20.41
II OHSW Alone 13.92
I Arresters Top 2 Phase
Alternate Structures
No OHSW 14.68
II Arresters Lower Phase
Alternate Structures –
Plus OHSW 9.21
III Arresters All Phases
Every Third Structure 18.16
III Arresters All Phases
Alternate Structures 15.26
III Arresters Top Phases
Every Structure 10.35
III Arresters All Phases
Every Structure 0
=====================================================================
NOTE: Arresters on the top phase of every structure results in a 26% improvement over shield wire alone. Therefore, for new construction the elimination of the shield wire and use of top phase arresters results in a better performing line.
Reference:

http://www.hubbellpowersystems.com/powertest/literature_library/pdfs4lib/OB/EU1292.pdf