Thanks MMcKenna, davelnr & others,
The required direction for is in a line that is parallel to the hand-rail. There are 2 Yagis. The one in the foreground is almost bang-in-line with the massive red pillar you see in the photo. This antenna in the foreground does not work at all. Both need to work reliably as part of a redundant pair. One at a time though.
The other antenna which is farther away, though not in line with the pillar has problems. Please observe that it is not exactly parallel to the required LOS, but rather turned away outwards.
This is because it NEEDED to be turned AWAY from the LOS. When aligned on the LOS - by sighting through the central beam- we failed to receive any signal. I thought that Yagis gave the best signal exactly along the LOS. However, with the massive iron structures around I am proved wrong!
Only when put in this direction, (approx 15 degs.outward of the LOS,) set by trial and error using walkie-talkies at both ends, did we obtain the best signal on the other end. The best RSSI we get at the buoy (5 kms. away) is only a measly -112 dBm with about 20% loss of data packets! This simply won't do. We need a stronger signal - RSSI > -75 dBm with no loss of data packets for a reliable link.
If I go for an additional 100 feet of LMR-400, The loss in the co-axial will no doubt increase to >3dBm. The 5W TX will deliver 37dBm with 33 dBm at the antenna. Let us now look at the Free-space loss.
Propagation loss in air for 156MHz & 5 Kms. =32.5 + 20*LOG10(156.0)+ 20*LOG10(5.0) = -90.3dB
Add the Coax cable losses of 3dB. So we get total transmitter power loss of 90.3 + 3 dB = -93 dB
Tx Power RTU at 1 watt => Transmit power of 30dBm (Assume lower power of 1 W and not 5W)
At the receiver, the received signal will be.. 30 - 93.3 = - 63dB. This means even at 1 watt transmitter power, the RSSI at 5Kms should be very strong..at -63 dBm. By this reckoning, since the required RSSI for reliable communication is as low as - 75dBm, we would still have about 12dB to spare! -Plenty-. This is to the coax-cable loss (in LMR 400 @1.6dB/100') for 7,500 feet!.
Note that I have not added the antenna gain of 10.5dBi for the Yagi and 3.0dBi for the Omni at the other end. Let this be set aside for the fade margin.
Coming back to the "Location" issue, the RSSI I get observe is only -112 dB. The coax + LOS path loss at present can account for only - 93dB. So the additional loss of about 17dB seems to come from the location. This causes me to think that even with a very long (200'-300') LMR we will be better off without the bad influence of the interfering structures.
I am in automation, not a radio engineer. But this telemetry link needs to be good. The formulas I used above, are from the Internet, where I seek answers for things I don't quite understand. Really appreciate all help. Am I right in thinking along above lines?