Wind/Solar Collection System, CDEGS FCDIST Calculation

[MAnton]

Assume a substation with one transmission source and one collector feeder. The transmission source has shield wires which are connected on both ends, and the collector system is UG, but also connected to the substation ground grid.

I am attempting to accurately calculate and understand the amount of fault current that will flow on the shield wire return path to the source substation, and am having trouble with what my "Central Site" ground impedance should be in the CDEGS FCDIST calculation.

Connecting the substation to the collector system grounds is necessary for safety, but I am wondering about how it effects the FCDIST calculation. Including the collector system ground network reduces the grid resistance significantly because of all of the additional buried conductor, but if I include this reduction in grid resistance to my model then the amount of current flowing on the shield wire is reduced (increasing the amount of current flowing to earth through my substation & collector ground grid).

The CDEGS Technical seminar reading guide (I attended this year) says to use "the ground resistance of the substation (when disconnected from the distribution circuit)" . . . but I am still not 100% sure what to do or what the physics/logic is behind that.

It seems that connecting to a large underground collector network that lowers the effective resistance of my substation grid should most certainly affect the amount of fault current flowing back to the source on the shield wire, but the above statement from the tech. seminar guide seems to suggest it doesn't?

Any help or thoughts here?

 

[jghrist]

If the collector is a significant distance from the substation, then I would treat it as a separate terminal. You would have two terminals (source substation and collector) plus the central station. The central substation ground impedance would be the MALT calculated grid resistance. Some of the return current will flow through the transmission shield, some through the substation ground grid and earth, and some through the collector cable shields to the ground at the collector.

 

[MAnton]

Hmmm ... I'll look into modeling it that way tomorrow.

There isn't much in the way of fault current coming from the collector system for my worst case fault scenario, will any fault current return on that path if it didn't originate from there?

My collector system ground network consists of buried phase conductors and a bare ground buried in a trench with the bundled phase conductors. I'm in a situation with extremely high fault current from transmission source, where I am very close to the transmission source substation. I'm struggling with the fact that connecting to collector system ground network takes my calculated grid impedance from ~1.5 ohms to ~.2 ohms. This sounds great, but if I run FCDIST with a grid resistance of 1.5 ohms, and then with a grid resistance of .2 ohms the amount of current going into the earth at my substation is hugely different. Lowering grid impedance just keeps increasing the amount of fault current flowing to ground at my substation (because less travels back on the transmission shield) which makes it hard to converge on a reasonable design.

Am I looking at this the wrong way, or did I just hit the lottery for bad design factors?

 

[MAnton]

Since there is minimal fault current coming from the solar collector system, I have decided to treat the project substation and the solar collector system grounds as a single element (similar to a substation with extensive counterpoise).

The effect of this is to lower the effective resistance of my grid, which decreases the amount of calculated current returning on the shield wire (using FCDIST) and therefore increases the amount of current flowing to the earth through my substation/collector system grid. Feel free to poke some holes in my assumptions here.

 

[jghrist]

There isn't much in the way of fault current coming from the collector system for my worst case fault scenario, will any fault current return on that path if it didn't originate from there?

Yes, fault current will flow through the cable shield through any grounds along the cable and through the collector ground back to the source substation.

Since there is minimal fault current coming from the solar collector system, I have decided to treat the project substation and the solar collector system grounds as a single element (similar to a substation with extensive counterpoise).

If you use MALT instead of the frequency dependent MALZ, the collector grounds will all be treated as being at the same potential. The impedance of the cable shield will not be taken into consideration. This will make the calculated current flowing in the collector system too high and the overall grid resistance too low. I think that it would be more accurate to make some assumptions about the collector system to treat it as a remote terminal. Not a lot of fault current will flow through the cable shield if the length is significant because of the cable shield impedance. You could model it both ways and see which is the most conservative in calculated touch-voltages.

 

[MAnton]

I am using MALZ. However, I'll make some average assumptions about the collector system and model it as a remote terminal to see what is more conservative & check back here to let any curious onlookers know.

 

[MAnton]

Modeling as a single substation with resistance of ~.3 yields a more conservative result (more current flowing to earth). But I may be modeling the underground collector system wrong in FCDIST.

The collector system has a bare ground wire buried in trench and extends about 1 mile radially from substation with solar collectors going off in different directions. I modeled as an above ground 4/0 line with the "collector system" as a remote source providing only 100 amps of fault current. I placed the "collector system" terminal at a distance of 3500 feet from my substation 10 "towers" and spans of 350 feet with ground impedance of 100000 ohms at each tower. The collector system impedance is what it was calculated to be in MALZ, and the central station impedance is also caluclated in MALZ with no collector system attached.

This was to simulate what I think to be a worst case where fault current has to travel far away back to the "collector system" terminal without dissipating anything in between. I also modeled a few different scenarios to try and simulate the fact that current will be leaking out along the whole 3500 feet back (because it is actually in the ground).

Results were that a significant portion of the fault current travels back to main transmission source. Another large portion goes to the collector source (much more than the 100 amps it is supplying). Very little (Comparatively) flows into ground at my actual substation. When I increased the number of towers back to my collector system source and decreased their resistance (to simulate the cable being in the ground) even less total earth current was seen at my central substation.

Seems logical, not sure if I actually trust the results.