Topband: W8JI verticals proximity to fence

[Lee K7TJR]

     Charlie, you did not mention what kind of an array you were planning. The answer would be very different with a simple cardioid array or 4-square as compared to a wide spaced BSEF type array. You also did not mention the length of the wire fence and if it was insulated and possibly close to some resonance near the operating frequency. 
    The higher the RDF of an array which most times comes right along with medium to high Front to back, the more sensitive the array will be to nearby objects. I did some extensive testing last year in an effort to quantify this very question. Frankly there are too many variables to provide an answer much more than a guess. What I have found is that if you can simulate the nearby wires or even objects in Eznec and you can get a feel for your particular situation. You might also consider your transmit antenna as a source of unwanted coupling as well. One of my best high performance arrays can "see" affects from my TX vertical when modeled at over 700 feet away. It has to be detuned during RX. 
     I think you can still download W8JIs rxvrhat eznec 20 foot element model on his site. I also think to get a reasonable answer to your question you would have to model the entire array however and then place objects such as fence wires and possibly use a slightly elevated radial near one element to simulate the buried element. Buried radials do couple to other objects in RX arrays.
    Aside from all that my gut reaction from all my work last year is that if you are building the 8 element circle array you are going to see a significant problem and it is going to come from the nearby radial coupling and likely not the top loading wire unless it is very parallel to the fence. If the fence is grounded every few feet I think the affect will be much less and will then manifest itself as a slight skewing of the pattern and slight reduction of RDF due to some extra signal coupling and  signal phase shifting in the close elements received signal. Personally I would skew any radials away from or at least perpendicular to the fence on that nearby element. 
    I would really encourage you to try to model this situation. I finally modeled all the power lines, irrigation pipes, telephone wires, buildings, and etc for my RX situation. I was amazed as to all the things coupling to vertical arrays and even Beverages. It was also important to vary the length of the modeled objects like power lines as they mostly were of indeterminate length so I assumed whatever was average or even worst. My biggest surprise was the local power drop with ground rod that looked remarkably similar to an inverted L on 160.

     My 2 cents or .005 gallons        Lee  K7TJR Oregon