Erattic engine efficiency

[Ryland3210]

I installed a flow meter before taking a long distance cruise two weeks ago, from Newburgh, New York up the Hudson River, Champlain Canal, Lake Champlain and Richelieu River and return. Heading north on New York 10% ethanol gas, 89 octane, the boat consistently used 1.6 GPH at 3500 HPM, for 1.29 MPG. The boat is a 2003, 7,000 pound 2855 Bayliner Cierra. The engine serial number is OL687277. It's a 5.7 liter, 260 HP EFI Mercruiser, single engine Bravo III I/O.

After refueling in Vermont on 91 octane ethanol free gasoline, it used 1.3 GPH at 3500 RPM, for 1.6 mpg. The engine also had significantly more power. Heading back south back into NY state, I refueled back on 10% ethanol 89 octane. When we got back on plane, the engine oscillated between the 1.3 and 1.6 GPH randomly any where from a few seconds to maybe 10 seconds. The throttle was held still while the boat accelerated on less GPH and decelerated on the higher GPH. YES, I really mean it! More fuel consumption and less power. After a while in this mode, the engine settled down to the low efficiency mode for the 50 miles at 3600 RPM for more than two hours all the way back to my home port. The best economy I saw on NY gas was 1.35 MPG up on plane. Economy was as good as 1.88 on VT gas. All test runs on smooth water with negligible wind. Ground speed by GPS, with the pitot tube gauge reading about the same. GPM readings were to three digits, e.g. 13.2

Things I tried: The flame arrestor was cleaned out with no effect. I unplugged the MAP sensor and plugged it back in, in case the connection was bad. Same with the air temp sensor. Took all the slack out of the throttle control mechanism and temporarily tightened it up firmly to hold it in position. Whenever turning on the ignition, it would always make two beeps and start right up. No more beeps after that.

 

[JoLin]

Afraid I don't understand the issue. You got better efficiency with higher octane, and with a mix of low and high octane in the tank your fuel consumption was a little erratic. If I read you correctly, adjusting/tightening the throttle cable improved things a bit overall. Despite the variations you saw (which I don't see as all that significant), your fuel consumption figures overall are right about where they should be for your engine and the size/weight of your boat.

My 24' Chris Craft with TBI 5.7 and a duo-prop drive would get a little over 2 mpg overall. 27' Four Winns with twin TBI V-6's (single prop drives) ran about 1.6 mpg, and the current boat with twin V-6's and four barrel carbs gets 1.2 - 1.4 depending on conditions. Didn't have a Flo-scan on any of them, so the figures are 'ballpart' but pretty close.

Are you having an engine problem that I don't see?

 

[Ryland3210]

It's a matter of degree. In all my boats in the past, including Mercruiser 5.0 in a 190 Searay and twin 454's in a 340 Searay Sedan bridge, when I got up on plane and set the throttle, the engine power stayed constant. In this case, the random change from "normal", to poor power with poor efficiency was not only a new experience, but dramatic. For example, upon on plane, 3600 RPM, 25 mph consuming 14 GPH, for no apparent reason, engine slows down to 3350 RPM at 18GPH and would start to fall off plane. It required full throttle at times to get back to 3600. Then backed off on the throttle to maintain 3600. Sometimes that required 18 GPH, but then the engine would suddenly gain power and shoot up to 3900 before I could catch it, back off on the throttle back to 3600, where it would consume 14 GPH. This might happen at any time, or run steady for an hour.

Tightening up the throttle linkage pivot joints was to prevent any change in butterfly position to see if that was the cause. It wasn't.

 

[alldodge]

I think your just seeing a carb and fluctuations that come with them. the boat bounces along. Getting better fuel economy with non-E over E10 is what I would suspect. Like to see what would happen with a fuel injected engine

 

[Ryland3210]

It's an EFI two barrel Rochester Throttle body. 25% change is a lot. Has a knock sensor. My amateur theory one: the ECM reacts to engine knock by radically retarding spark. Theory two: Losing spark to a cylinder intermittently.

 

[alldodge]

Wow I really blew that one, first post said TBI :facepalm:
Suggest checking your fuel pressure, what is your engine serial number?

 

[Ryland3210]

I don't think you made any mistake. Maybe it's caused by the terminology as I understand it. The engine was sold as an EFI. To me, that means Electronic Fuel Injection. The fuel injection is controlled electronically, so it appears to be an EFI to me. To me, TBI means Throttle Body Injection, which it has. It's a Rochester two barrel with injectors over the inlets. The serial number is OL687277.

Please tell me what I have! On another post, there was a link to a site where I entered the serial number and it came up with the whole list. I haven't been able to locate it. I saved the actual site link, but it won't get me to the same great resource. The Mercruiser parts website won't get me to the same place, only to generic engine types. Could you help me find it again?

Thanks.

 

[alldodge]

You have an EFI engine which uses TBI. The other type of EFI engines is MPI (Multi Port Injection), all are EFI's. There is Mercruiser (OEM) and Mercruiserparts is a site the sells Merc parts. When entering your serial at Merc use a 0 (zero) not an O (letter) as the first digit
Mercruiser https://www.mercurymarine.com/en/us/parts-and-accessories/parts-catalog
Mercruiserparts http://www.mercruiserparts.com/SelectProductLine.asp

 

[Ryland3210]

Some background:

I had a 1989 340 Searay Sedan bridge with twin 454's with carbs. I had flowmeters on both engines then too, so I could tell as soon as anything went wrong. By the end of the first season, some of the plug wires failed. I was surprised, but the service tech told me it was a common problem, due to the 65,000 volt Mercruiser ignition system. It cost me $350 to replace them with the 7mm original equipment Mercruiser wires. I decided I could come up with something better. I bought some new wires aftermarket that were the first version of the silicon rubber spiral wound types now commonplace. Then I assembled them through 20mm OD, 7mm ID Silicon rubber hose with heat shrink to make the transition from hose to end caps. Never had a problem with plug wires or misfiring again.

I decided to start with the simple stuff. It's not in my nature to shotgun solutions without knowing the cause. However, I was going to change the plug wires anyway. I'm changing the high tension wires to 8.8 mm silicon rubber spiral wound graphite cores. I am not a fan of black injection molded, thermoset or Bakelite plastic distributor caps, but haven't been able to find a better alternative for this engine like the tan ones pioneered by Chrysler Corp in the sixties, so I bought an OEM Mercruiser replacement. If that doesn't do it, at least the wires will be an improvement in reliability.

 

[wahlejim]

Just thinking outside of the box here, have you checked your prop for damage to the blades, hub, ventilation caps? Also, what was the current on the river like?

 

[Volphin]

Sounds to me like the TPS is defective and is sending errant readings to the ECU. I think the Mercruiser part number for your engine is 853678T. However, the part is quite common for less $$. Try AutoZone and request a TPS for a 98 Chevy MPI? if memory serves.

 

[tkat1000]

being a gm auto tech for many years. servicing EFI systems, sound very well like a faulty throttle position sensor, giving erractic readings.which may not set a code. you'll need a scan tool to see it. one other thing to check would be a rotted vacuum line to the manifold pressure sensor. the rubber many have a hole or just soft enough to collapse under load..

 

[Ryland3210]

Thanks for the suggestions.
WahleJim: The problem occurs randomly regardless of river depth or current. I doubt the props are damaged.
Volphin: The engine is a Rochester Throttle body injection type. It is not an MPI. However, maybe the same TPS is used. Do you have another Chevy candidate using the Rochester Throttle Body? The part number for my engine is 805226A, listed at about $73.
tkat1000: It would be great if I could get the equivalent from the local Auto Zone. The manifold pressure sensor is attached directly to the manifold. No hose involved.

 

[Scott Danforth]

since ethanol has less BTU's per gallon, as well as most early EFI tunes dont handle the appropriate change to AFR when ethanol is introduced, you get a reduction in power output as ethanol is increased. (the reason why your VT fuel had more output) also, there is usually a bit of residual "stuff" in the fuel depending on where you fill up. potential cause of oscillating results in NY. my guess is when you got to NY, the fuel had more "stuff" in it. some of the "stuff" probably only affected burn rate, some of it probably affected your fuel flow meter operation.

during lab testing of diesel engines, I had a case where according to the fuel flow meters, the motor was producing fuel. turns out a small piece of dirt had lodged itself into one of the flow meter gears on the incomming meter and was causing it to turn slower resulting in an errant reading.

 

[Ryland3210]

Thanks. The oxygen content in ethanol can be considered partially burned hydrocarbons. The whole purpose of gasoline is to burn the carbon and hydrogen with oxygen to make energy.

Since this flow meter was installed right before this 700 mile cruise, we had the opportunity to check calibration with numerous fill ups. The meter has an adjustment to recalibrate, which we used with each fill up. It was off quite a bit for the first two fillups. Too low by a lot, then too high a few percent as we overshot the adjustment then it settled down. Generally, it stayed within 1-3 percent on a 50-65 gallon fill up thereafter. We didn't notice any substantial error, even with the 25 percent variation in consumption. Interestingly, Flo Scan does calibrate separately for diesel and gasoline. All other things being equal, I suppose it's the viscosity difference. Not certain, but I suspect it uses a turbine type sensor.

Anyway, my next move it to take it on a test run with the new 8.8mm Accel plug wires, then a new TPS. Not sure whether to have the engine scanned before or after the TPS. Whichever opportunity comes first! I'm tempted to remove the TPS, hook it up to a voltage supply, and look at the output voltage with a scope as I rotate it back and forth. Maybe it will provide a nice smooth signal, or maybe it will provide a "scratchy" signal. All potentiometers eventually provide erratic output in my experience, except possibly for conductive plastic types. Don't know what type the TPS is, but tempted to find out.

 

[Scott Danforth]

most likely it is providing a pulsed signal. basically its a turbine wheel and a proximity sensor. the sensor picks up the pulses of the blades as they rotate past.

some high volume flow meters use a series of orifices and measure pressure drop, however the low flow variety are either turbine wheel for the lower cost, and meshed gears for the higher cost units. both of these use a prox to pick up rotation pulses.

 

[Ryland3210]

I read in the manual that plug wires near the knock sensor can cause interference with its signal, so I used several wire separators and routed them away from the signal lead. The Mercruiser authorized technician had not bothered to use the existing wire supports on the starboard side where the knock sensor is. I was careful. With the new wires, distributor cap, and wire supports, the change was dramatic. RPM was rock solid whereever I set the throttle and the engine used about 15% less fuel.

Since I already ordered the TPS, I will try replacing that next and see if GPH changes. If it makes no difference, I'll have a spare on hand. Next is to test the knock sensor's output. It is only a single lead device. Anyone have any comments?

 

[Ryland3210]

All the engine sensors tested good, so I guess that's as good as it gets.