You may already have fixed the problem by disconnecting and reconnecting the ECU plugs. The water may have corroded some ECU terminals, leading to a poor connection for ECU terminals that ground the main relay/fuel pump circuit and that power and ground the fuel injectors.

G101 is the master ground for the main relay, ECU, and sensors.

 

Not quite. When I got to work I took the computer in with me to keep it cool. Brought it out, plugged it in, wouldn't start. I do see the G101 now but I'm away from my garage. Will have to come back for it tomorrow.

Found a salvage yard computer for $85. Will try my luck on that tomorrow and clean up the G101.

 

Replacing the ECU is worth trying, but the intermittent nature of the problem sounds more like a poor electrical connection. Definitely sand clean and tighten the G101 connection. Also carefully inspect the ECU connector terminals (both the ECU and connector sides).

 

I got it running just as my wife came to pick me up. Sometimes the fuel pump would kick on and sometimes it wouldn't but it never started and I always saw the same 16-light sequence from the ECU. Usually I wait for the relay to click and the pump to stop before trying to start but the time it started I turned it over when I heard the pump come on.

So anyway, I got it home. Ron, thanks for the description and the diagram. The G101 wasn't corroded at all but I took a wire brush to it, cleaned it up, wiped it off and put some dielectric on it to be safe.

My new ignition switch came in today and I installed it. I took the old one apart and there is some pretty good pitting on both of the battery contacts.

The car has been starting fine since I got it home so it's hard to know if the new switch and/or cleaning up the G101 helped. Ron, the pins and connectors to the ECU look to be without corrosion. I talked with my Dad and he seemed to think the fuel pump could be going bad. How does that sound to you? My goal for tomorrow is to try and get it to not start in my driveway to I can test for spark.

I'm hoping all of these details can help someone else in the future (provided this thread doesn't end with, "Gave up, bought a new car."). Thanks for all the help!

 

A bad fuel pump would not throw code 16. Neither would a bad ignition switch. For code 16, you generally focus on the main relay, ECU, and fuel injector power and ground circuits.

 

The CEL would still do the 2 second cycle with a bad fuel pump. Just no fuel would flow. I think you said your CEL stays on, does not cycle when it won't start. If that's the case it's something else.

 

Actually, mk378, the CEL does the 2 second cycle no matter what the outcome. Sometimes the fuel pump kicks on sometimes it doesn't but the CEL always cycles. It never stays on.

 

Let the car sit in the sun all day yesterday. High 80's here in Buffalo. I've got a thermometer in the car. Cabin temp was about 121 F. Had no issues starting the car all day. Had it sitting in the sun again today and it would not start around 1. Pulled a plug and tested for spark. There was spark. The plugs were all a little heat-worn looking and 3 of them had a pretty good amount of oil on them (the engine loses about 1 qt between changes. has been that way since I got it). I had put in new plugs, wires, cap, rotor last August, about 12k miles ago. Ran out to the parts store, picked up a new set, came back, gapped em (1.1mm), put them in and the car started fine. But there's too many variables going on (windows and hood were open for about 30 min while I picked up the parts). I closed everything back up and sort of hoping it won't start later so I can check for fuel. The real test will be tomorrow at work. The cabin temp, with windows cracked, got up 136F after sitting in the parking lot for just a couple hours around 1pm the other day.

 

Wouldn't start again after sitting in the sun (110F cabin temp). No fuel pump hum, no relay click. Tried 3-4 times. Banged on the fuel tank with a wrench; hum, click, started up. I'm leaning towards wiring or the pump itself. Any thoughts?

I'll pull out the backseat in the next day or two to test the wiring the next time it won't start.

 

So the CEL cycles, but relay doesn't click? That's unusual.

If the relay doesn't click there's no way the pump can run.