Replumbing an Outboard for Heat

There are cast in passages, installed passages, and voids made during casting. Routes are particular to individual engines, and manufacturers. If I ever do this, I'll have to disassemble the engines in question, at least to the level of removing the lower units so that I can see, for sure. But on the engine in the image today, we might, for example, drop in a tall cup that ¾ fills the drainage passage, with the take-off tapped through the side of the LU at the bottom end of the cup. For the cup, imagine an 8", 20cm 1½" 4cm tube. Hot water splashing down the passage splashes into and fills the cup, or just falls past it if it misses, or the cup overflows. Consider the size of the hot water passages at 8 vs the size of the passages between them. That water is just splashing around as it drops out the bottom.

 

I've had a consult with a heating engineer. He says the 130kW available from a pair of 150s is enough to heat an office building. But the water coming out at 160-200° F, 70-90° is too hot. Unless the flow has very low turbulence it's too likely to get local cavitation occasionally, and pit through a simple heat exchanger. Better to use a mixer and get the temp down to 50/120. We are discussing, as a casual thing over coffee, materials for a heat exchanger that can handle extremely hot sea water.

Some engineering assumptions: water enters the system at 10°, exits at 55°c. Heat rejection is half by water, half by hot exhaust:. 30kW by water. Given that delta T, flow is 20l/minute.
Translation: intake temp is 50°, outflow is 130°, flow rate is 5 gallons a minute.
Naturally these assumptions are only adequate for a first feasibility discussion.

@ondarvr , why would increased cooling requirements be met with reduced cooling flow? Centrifugal pump output being a square function of rpm, where power and therefore heat rejection demands are roughly linear?

 

The water will not be in the 160-200F range when the motors are running at anything much above an idle. The bypass valves open up and raw water will bypass the thermostats. The real temperature will fluctuate a great deal depending on ambient water temperature.

 

Well, as I said, we don't want water that hot anyway. But how hot, what rate of flow. Ambient temperature doesn't change much. Thermal mass of the Pacific Ocean is significant.

 

If you have ever put your hand under the telltale; it cools off a lot at high speeds.

Variations in ambient temp, engine temp, flow rates are gonna make that a real challenge.

Aquastats are used in conventional systems.

 

As I've said many times, the telltale is before the block, not after.

 

Sorry to be grumpy about it, but the telltale is shortly after the impellor. The water has run up the leg, and picked up some heat from the exhaust on its way out in a parallel passage, but not much, because carrying heat back into the engine again really defeats the purpose of a cooling system.
Just at the top of the leg, before the water enters the block, a little bit of it squirts out the telltale, just to show you the impellor is working. You can't tell anything about what is going on inside by water from the telltale.

If @ondarvr 's comment about bipass water mixing with the block water is correct, there must be a really substantial flow of water from the impellor, because half the heat rejected from a 150 at wot, with the other half in the exhaust is a bathtub full of water, hot enough to kill you, every 6 minutes. I don't know what percentage is actually carried away by the exhaust, but water has 3500 times more ability to carry heat than exhaust does, per unit volume.

 

Assuming the 150 is reasonably efficient, it's taking about 105 cubic feet of air a minute, vs about ¾ cubic feet of cooling water. Imperial units aside, the ratio is 140 to 1, air vs water. With a specific heat 3500 times higher, the delta t for the exhaust would have to be 25 time higher than the cooling water. Or 1125°c, or 2000° Fahrenheit. Hmm, that doesn't seem right. Too many assumptions there. But clearly, waste heat from the engine is largely in the water.

Can one of the engineers check my assumptions? @gonzo , @Barry @baeckmo ?

Aw heck. Ideal gas law. I forgot that hot gases expand. 105cfm in, several times that out.

 

You can overheat even small a motor by running it in a drum of water, that's because it cycles through motor many times. Even a small increase in water temperature running through the motor each time adds up. Plus the thermostat keeps the motor at a higher temp at lower RPMs.

You'd need to measure the actual increase in temperature at higher RPMs on a single pass through the motor.

The water temperature can fluctuate a bit depending on what part of the West Coast you're in, but if it's 60F, and you get a 20F rise, is that enough?

 

Oh yes, that's plenty. The flow would be massive though. I definitely need to move, finish Serenity, build a suitable test stand, acquire a suitable engine, tap it for instrumentation, maybe learn some coding so I can write some simple code, design and execute a test program....

But a 20°f rise would require a flow of 20 gallons per minute. 75 liters a minute, roughly speaking. A massive demand on the impellor. And seems unlikely. Particularly when the block wants to stay between 160 and 200° F.

 

The block water does not want to stay at 160F, it will drop significantly once the bypass opens.

The 160 thermostat is to ensure that it runs at low RPMs correctly and doesn't foul plugs.

 

Just to offer some help. The aquastats on the Webasto diesel system don't function below 120F because the system would cool off too fast otherwise.

 

Took another look at OB impellors. They are positive displacement pumps, not centrifugal pumps, so flow rate is directly proportional to rpm. Presumably a safety margin would require more capacity than peak cooling needs, so there'd always be some water bypassing the block. Water leaving the block will be at whatever temp the engineers wanted it to run at - 160 to200F, but bypass ratio would govern the temp of available water. I'm about at the limit of what I can determine from first principals, physics and understanding of good engineering practices. In the absence of more data, I don't have anything else to say.

 

Outboard cooling systems are rather crude, and error on the side of not allowing the motor to overheat. Since overheating is the primary concern, excess cooling capacity designed in.

Although I own Mercs now, most of the work I've done was on Johnson/Evinrudes, which had 140F thermostats, I've never checked what's in my Mercs.

You can use a non contact thermometer and check the waterjacket temperature, it's fairly accurate for discharge temperature of the water. This needs to be done on the water at the RPM and load you would expect to normally run at. Old school used a wax pencil that melted at a specific temperature.

For low RPM testing of the thermostats it can be done on a garden hose, this doesn't work for higher RPM temperatures.

This will give you a ballpark range of what to expect. After the loss of temperature through the plumbing, heat exchangers, etc, you may be able to calculate whether there is enough water at a high enough temperature to heat the desired space to the extent that it will make a real difference.

You only have access to 50% of the total output of the waterpump. And any restrictions you put in place by diverting the flow will reduce/slow the amount traveling through that half of the block. This will increase the temperature on that side compared to the other side. There is normally a slight difference between the two, but you don't want to increase it.

 

Take the tell tale from your engine, run it into the boat and keep your hands warm with that but the telltale is in different places on different engines so it might be hot and it might not.
99% of outboards have 143f thermos