Glow Plug Solenoid Failure

Now I have the resolution of the problem I found on 7/6/12.
When I turned off the battery selector switch (BSS) the shore power charge finally recharged the batteries.  I kept looking, and I observed that when I started the engine, its generator was making 50 amps, but the battery monitor was seeing only 5 amps.  I had never noticed this before, but it seemed quite suspicious.
The only load of such a size on my boat is the glow plugs, and I found that they were energized all the time while the engine was running.  The glow plug solenoid (GS), it turns out, was burned out, and it it remained in the closed position, even though the energizing signal was removed.
Checking my records, I learned that the GS was replaced this spring, as part of a starting circuit troubleshooting procedure.  They also replaced the corroded glow plug power bus and one or two plugs.  Naturally, I called the mechanic and requested that the shipyard replace the defective part.
Mystic Shipyard consented, but suggested that some other defect might be behind both failures.
When the new GS arrived, I installed it, and I found that the GS was still energized all the time the engine was on, although it would turn off if I removed the wire from the "S" terminal; the defective one would not.
To investigate the problem further, I reviewed the circuit schematic that first came with the engine.  I found that Westerbeke connected a wire to the "I" terminal on the GS which was connected to the "S" terminal on my engine.  I checked my own documentation of the engine and other boat wiring, and saw that the wire had been connected this way for many years.  Many mechanics, as well as I, had worked on various engine problems, but none had spotted this anomaly.  Reviewing my records even more carefully, I saw that the GS had been replaced at one shipyard or another about every second year.
I called Hansen Marine, and Bob Hansen confirmed that the green wire between the 10A circuit breaker and  the "S" terminal should have been connected to the "I" terminal.
As designed, while the engine is being started, the the "I" terminal feeds the green wire, which feeds the lift pump and fuel solenoid through a 10A self-resetting circuit breaker.  The "I" terminal receives power through the GS, when the GS is energized by the "WARM" switch in the starting sequence. When the oil pressure (OP) switch closes, the lift pump and fuel solenoid receive power through the contacts of the OP, so they will shut off if the engine loses oil.  The "I" terminal is not closed when the "S" terminal is not energized, so no current passes through the 10 A circuit breaker.  
If the "S" terminal is connected to the green wire, the OP passes power backward though the 10A circuit breaker, so it latches the GS energized.  This was the source of my problem.

According to Hansen, there are many generic GS models in the aftermarket, some of which do not have the "I" terminal.   He thought that someone in the dim past did not understand its purpose, and bypassed it to make the engine go.
I went though this elaborate discussion, thinking that someone else may not have understood how this popular engine is designed, and may have the same error causing repeated GS failures.
Hansen also noted that a pull-down resistor has been added on terminal "I" in later versions of this diagram to prevent spurious alarms from the OP switch; I will do so if the problem occurs.