Distribution Line GPR

[jghrist]

I am analyzing the grounding of a 911 center served from a 12 kV distribution line. I have calculated the Ground Potential Rise (GPR) of the ground system for a 12 kV fault at the transformer pole serving the center. I am using SES MALT and FCDIST software. MALT calculates the ground resistance and surface potentials. FCDIST determines how much of the fault flows back to the substation through the earth and how much flows through the neutral.

The GPR at the center is high, but not unreasonable, about 1460V. I decided to see what would happen at the 911 center if there were a fault on the transmission side of the 100-12kV substation serving the 911 center line. I didn't expect this to be a governing factor, but it turns out that the GPR at distribution line pole grounds and at the 911 center are almost as high as the substation GPR, over 11 kV. This is essentially a transferred potential problem, with the distribution neutral propogating the substation GPR to all of the distribution pole grounds, with little attenuation.

If the GPR is this high at a distribution pole where there is no ground grid, just a single rod, the touch voltage would be dangerously high. We use IEEE Std 80 and go to great lengths to reduce touch voltages inside substations to safe levels. There are no standards in the US that I know of for touch voltages at distribution poles. Does anyone know of any studies made of this situation or practices to alleviate the problem?

 

[alehman]

It seems a little odd to me that there would be enough ground fault current available on the 100kV system to create that much GPR. How does it compare to the 12kV fault current? As you metioned, I would not expect it to be a factor.

Possibly silly question - is the 100kV system solidly grounded? Have you discussed your concerns with the utility company?

 

[jghrist]

The 100 kV 1Ø fault level is 6654A (obtained from utility). 2235A flows through the substation ground grid and the rest through the shield wires of two incoming 100 kV lines and the neutrals of two outgoing 12 kV feeders. The substation grid resistance is 5.158 ohm (measurement provided by utility, and within expected range considering soil resistivity). The 100 kV is solidly grounded.

We measured the soil resistivity near the substation and modelled it with SES RESAP program to get a two-layer model with an 8.1' upper layer of 1456 ohm-m and a lower layer of 710 ohm-m. Soil model at the 911 center was two-layer with a 1.4' upper layer of 290 ohm-m and a lower layer of 1943 ohm-m.

An 8' ground rod in the 911 center 2-layer soil would have a resistance of 527 ohms. This is what we used for 12 kV feeder structure ground resistance. We used an average resistivity of 1220 ohm-m in the FDIST program because this would give the same 527 ohm resistance for an 8' ground rod.

12 kV 1Ø fault current is 6172A at the substation and 2920A at the 911 center.

 

[stevenal]

jghrist,

You seem to be assuming the high GPR leads to directly to high touch potentials. It's the GPR gradient that governs the touch potentials. A high GPR fairly evenly distributed over a multi-grounded neutral distribution system sounds like a low GPR gradient to me. Did you model the gradient?

 

[jghrist]

stevenal,

FCDIST calculates the structure GPR at every pole ground on the feeder. If there were no current and infinite pole ground resistances, then the voltage of the neutral would be at the station GPR for its entire length (no voltage drop in the neutral for no current). There is some current so there is some voltage decrease along the feeder, but in my case it is not much.

On the 911 center feeder, the neutral current is 770A at the substation and 360A at the end. Structure GPR is 11.5 kV at the first pole ground and 11.3 kV at the end. The ground resistance at the 911 center is 31.6 ohms. All of the 360A neutral current in the last span goes into the 31.6 ohm ground, causing an 11.3 kV GPR at the 911 center.

As you say, the touch voltage is different from the structure GPR, but the gradient around a single ground rod is pretty high and the 11.3 kV GPR gives rise to a touch voltage of about 6.5 kV.

 

[stevenal]

jghrist,

A gradient of 11.5-11.3=0.2 kV along the entire feeder? And a 6.5kV gradient within a few feet of an individual pole ground? Doesn't pass the reasonableness test with me.

You might want to check with SES. I understand their support is pretty good.

Our practice is to put a ring around pad mounted transformers and other pad-mounted metallic equipment. No standard involved that I know of, other than our own.

 

[jghrist]

The 6.5 kV gradient is between the ground at a pole and the earth surface 3 feet from the pole. This is not the same as the voltage gradient along the metallic neutral. The voltage gradient along the metallic neutral is caused by current flowing in the neutral and the earth surface gradient is caused by current flowing in the earth from the ground rod.

I have discussed this with SES. They agree that this could be a danger but that there are no procedures to protect people in this case, at least in North America. They suggested that the problem is most severe in my case because of the high soil resistivity and ground rod resistance.

I tried changing the soil resistivity to 100 ohm-m and the pole grounds to 50 ohms. The substation GPR went down to 4.4 kV, but there was still not much reduction along the 12 kV neutrals. The structure GPR at the end of the feeder was 91% of the substation GPR.

The ring around the padmounted equipment sounds like a good idea, but there are a lot of pole grounds and grounded guy wires out there that don't have any voltage gradient reduction. Thing is, I have looked at the touch potential on padmounted equipment before for faults at the equipment. There generally is not too much of a problem because most of the fault current flows back through the cable neutral instead of through the earth and there is a lot of voltage drop in the neutral so the GPR drops off quickly as you get away from the fault location. But for a transmission line fault at the substation, there is not much current in the neutral to reduce the GPR along the line.

 

[CastleHydro]

Your findings seem reasonable, although somewhat alarming! It appears that the utility's available fault current is very high, and combined with the poor soil resistivity is leading to a dangerous condition. Under the circumstances, it is unlikely that the quality of the substation ground could be improved much, but I wonder if the utility should consider current limiting reactors. This transmission line seems pretty beefy for feeding two 12kV distribution lines!

 

[jbartos]

Suggestion: Follow IEEE Std 80-2000 IEEE Guide to Safety in AC Substation Grounding, Annex D "Simplified Step and Mesh Equation"