An interesting thing I think is to look at graphs of tether reel in and reel out speeds, or rotation speeds for torque transmission. If you have a sense of what those look like you could look into if connecting multiple ground stations to a single generator or adding flywheels or something like a continuously variable transmission (CVT) for example could make sense.
In conventional wind you can’t really connect multiple wind turbines together like that because the rotor goes around the tower. And a flywheel also makes less sense there because the tower would need to support it, the rotor is already a flywheel and power generation is mostly continuous. All of that isn’t the case in the systems this topic focuses on.
Assuming the ground stations are in a hexagonal grid, the generator station is at the center of the center hexagon, and the transmission method is only allowed to go along the edges of the hexagon. So if you imagine the vertexes of the hexagons as nodes in the power transmission network, then you could perhaps connect a single generator station to 6, 12, 24, 36 systems if you allow for 0, 1, 2, 3 relay stations let’s call them for now. Assuming 0.9 efficiency of the relay/transmission method then the power from the outermost system delivered to the generator station would be reduced to 0.9^4 = 0.6561 for 36 connected systems or 0.9^3=0.729 for 24 systems. Actual efficiency could be higher.
There are of course other possible geometries, like a number of lines coming from the generator station that at specific points along their length have multiple other lines branching from them. Perhaps lines coming from nodes further away from the generator station don’t need to connect to all the nodes along the path, instead just running parallel to the lines coming from nodes closer to the generator station.
The possibility of only needing one generator station for 24 systems for example is interesting enough to warrant further exploration I think. You could also explore replacing the generator in the station with something else, say 6 smaller generators, or a single smaller generator as a pilot study.
Simpler would be to house all the drums in the generator station also and connect them to the generators, with the tether from each system going all the way to the generator station, the other nodes just redirecting the tether. Something like this was also something Zhonglu High Altitude Wind Power Technology - 中路高空风力发电技术 did or planned to do, 4 systems for one generator station.
The nodes could for example look like a mix between water towers and electrical transmission towers. The tethers or rope drives would then be raised from the ground, like suspended transmission lines are. The topmost sheaves could be for the ground stations surrounding this node and could go around an upright, vertical axis, static, cylinder, to reduce their inertia. The individual sheaves would probably have their own bearings to allow for independent rotation. The cylinders for lines coming from nodes further out could have a vertical or horizontal axis I think, perhaps depending on if the nodes are in line or not.
Assuming you have a number of tethers or rope drives coming from a node, in the generator station you could place the drums for them in line with the tethers, one behind the other, their axes perpendicular to the lines: [] [] [] [] [] [] — [drums: [], generator shaft: ), or slightly staggered to stagger the transmission from the drums to the generator shaft.
Additional tethers from additional nodes could then come into the generator station parallel to the other tethers and go to drums that connect to the same or different generator shafts, resulting in a (staggered) grid of drums.
The transmission from the drums to the shaft could be something like a CVT that connects to one of two gears, depending on if the line is reeled in or out. Like this there would still be a sudden direction change of the drums when going from reel out to reel in, so you’d need to find a solution for that, in the likely case that is a problem.
One wonders if it is a problem if the kites don’t pull exactly the same. I think perhaps not, just like it isn’t much of a problem if reindeer don’t pull exactly the same on a sleigh. The CVT should also optimize reel out speed for an individual kite (train).
Instead of the tethers / rope drives coming out of the generator station in a straight line and continuing on in a straight line, you could bend them a little at each node, creating spirals. You’d potentially minimize tether bending while still evenly distributing the ground stations around the generator station, minimizing the number of nodes, and optimizing rope drive span.
On a ship, or a movable platform, perhaps you’d put the generator station at the back of the ship.
Simpler but not better. Better would probably be to try to avoid wear on the expensive slow moving tether, if you’re doing bounding / pumping AWE, as much as possible and have a drum on each ground station instead. You can then connect the ground station to the closest node with a rope drive. You’d increase the speed of the rope drive, allowing you to use a thinner and cheaper tether and replacing that instead more often.
This system of nodes then also becomes an option in all places where you want to harvest rotary motion, making it more interesting to look at.