Additional Engine cooling

[wergie]

OK you lot, pay attention please..

I'm about to replace my rather knackered old engine with something a little more powerful.

My current engine is rated at 16 hp, and has a vertical skin tank has been sized accordingly.  It's replacement will be 30hp, hence the need for some additional cooling.

Now, my cunning plan is to build a horizontal keel tank in the drip tray under the engine. The tray is about 6' x 2', so there's plenty of area, but what I need to know is Will It Work.

Opinions so far seem to be divided...

If anyone has any experience of horizontal cooling tanks, if they work and how they are baffled internally I would be delighted to hear from you.

 

Steve

[Jen-in-Wellies]

What is the thickness of the base plate? What is the thickness of the side plates? What is the area of the existing side skin tank? Heat flow will be proportional to the thickness of the steel plate, so a thicker plate will need correspondingly bigger area I reckon.

Folks with more knowledge than me about such things here.

 

Jen

[RLWP]

We have a horizontal skin tank, it is the bottom of the bilge around the engine and so is a horseshoe shape

 

A Hudson I was working on this week has a rectangular tank as you described, between the engine bearers

 

Richard

[john6767]

3 minutes ago, RLWP said:

We have a horizontal skin tank, it is the bottom of the bilge around the engine and so is a horseshoe shape

 

A Hudson I was working on this week has a rectangular tank as you described, between the engine bearers

 

Richard

Don’t all Hudson’s have the skin tank on the base plate, so I guess it works, but possibly more difficult the bleed.

[RLWP]

11 minutes ago, john6767 said:

possibly more difficult the bleed.

 

Never had a problem bleeding ours, it is however hard to empty. I notice that our tank has several bolt heads for bleed screws

 

Richard

[ditchcrawler]

You could always go well over the top with area and baffles.

[Boater Sam]

Another external tank on the swim? Cover up some of those scrapes and rust pits?

 

To be considered is the total water volume. If it gets too big, you have the problem of containing the expansion. I had this problem because my skin tank is in the triangle of the port swim. so it is very thin at the front end but very wide at the rear end. The expansion exceeds the volume of the exhaust heat exchanger header tank.

So when its full it dumps water, then when cold there is insufficient water for circulation to occur leading to flash over heating. I have an additional expansion tank.

[john6767]

29 minutes ago, RLWP said:

 

Never had a problem bleeding ours, it is however hard to empty. I notice that our tank has several bolt heads for bleed screws

 

Richard

Makes sense!

[alan_fincher]

53 minutes ago, Jen-in-Wellies said:

What is the thickness of the base plate? What is the thickness of the side plates? What is the area of the existing side skin tank? Heat flow will be proportional to the thickness of the steel plate, so a thicker plate will need correspondingly bigger area I reckon.

I don't reckon it will make a lot of difference.

Steel transfers heat pretty well.

I haven't done the physics / arithmetic, though.

 

I'd also say that a more powerful engine will only dissipate more heat if you try to use the extra power.  If the old engine pushed along at full canal speed and didn't overheat, then if you travel at the same speed with the new one, it may well also be OK.  Of course if you need more power because the old one didn't go fast enough, or because you need to push against rivers you probably will need a tank with overall greater surface area.

Edited August 1, 2018 by alan_fincher

[Robbo]

Another option would be to have some raw water cooling via heat exchanger.   When you need a little more cooling  hit a button to turn a pump on.   You could also add a heat/exchanger for your central heating and warm up ya boat instead.

[Jen-in-Wellies]

24 minutes ago, alan_fincher said:

I don't reckon it will make a lot of difference.

Steel transfers heat pretty well.

I haven't done the physics / arithmetic, though.

 

I'd also say that a more powerful engine will only dissipate more heat if you try to use the extra power.  If the old engine pushed along at full canal speed and didn't overheat, then if you travel at the same speed with the new one, it may well also be OK.  Of course if you need more power because the old one didn't go fast enough, or because you need to push against rivers you probably will need a tank with overall greater surface area.

Hi Alan,

The physics says that heat transfer (watts) is directly proportional to area, inversely proportional to the thickness and directly proportional to the temperature difference between the two sides, the difference between the coolant and canal temperatures. Doubling the steel plate thickness would need a doubling of the skin tank area to get the same heat transfer and cooling effect. Overplated boats can suffer cooling problems where they never did before for this reason.

 

There is a recent post from Tony Brooks that suggests 1sqft per 4HP of engine power as a good rule of thumb. I am guessing that this is on a typical 6mm side mounted skin tank and I reckon you would want to derate that to 1.6sqft per 4HP for say a 10mm baseplate. This would suggest the 30hp engine is going to need 7.5sqft of cooling at 1sqft per 4hp, or 12sqft when derated to 1.6sqft per 4HP. By nice coincidence, the 6'x2' baseplate skin tank the OP is proposing is 12sqft! 

 

I agree that a 30hp engine isn't going to be as stressed, or run anything like as near to its limit as a 16hp one when canal cruising. The exception, as you say, is if it is ever used on long and perhaps tidal rivers. I cruise the tidal Trent quiet often and have come across narrowboats that are overheating. The owners have commented that it has never happened before. The tidal Trent is notorious for this. Constant running for hours on end at higher than normal revs.

 

Jen

[RLWP]

16 minutes ago, Jen-in-Wellies said:

Overplated boats can suffer cooling problems where they never did before for this reason.

I disagree. The reason they have problems is because the two skins are not joined together, so the thermal transfer between them is poor.

 

Richard

[Jen-in-Wellies]

1 minute ago, RLWP said:

I disagree. The reason they have problems is because the two skins are not joined together, so the thermal transfer between them is poor.

 

Richard

Good point. Didn't think of that. The discontinuity between the two layers of steel will add an extra barrier, especially as the old hull outer is going to be a bit pitted and rough.

[RLWP]

4 minutes ago, Jen-in-Wellies said:

Good point. Didn't think of that. The discontinuity between the two layers of steel will add an extra barrier, especially as the old hull outer is going to be a bit pitted and rough.

And as soon as you get a small air gap, the conductivity becomes poor

[dmr]

19 minutes ago, Jen-in-Wellies said:

Hi Alan,

The physics says that heat transfer (watts) is directly proportional to area, inversely proportional to the thickness and directly proportional to the temperature difference between the two sides, the difference between the coolant and canal temperatures. Doubling the steel plate thickness would need a doubling of the skin tank area to get the same heat transfer and cooling effect. Overplated boats can suffer cooling problems where they never did before for this reason.

 

There is a recent post from Tony Brooks that suggests 1sqft per 4HP of engine power as a good rule of thumb. I am guessing that this is on a typical 6mm side mounted skin tank and I reckon you would want to derate that to 1.6sqft per 4HP for say a 10mm baseplate. This would suggest the 30hp engine is going to need 7.5sqft of cooling at 1sqft per 4hp, or 12sqft when derated to 1.6sqft per 4HP. By nice coincidence, the 6'x2' baseplate skin tank the OP is proposing is 12sqft! 

 

I agree that a 30hp engine isn't going to be as stressed, or run anything like as near to its limit as a 16hp one when canal cruising. The exception, as you say, is if it is ever used on long and perhaps tidal rivers. I cruise the tidal Trent quiet often and have come across narrowboats that are overheating. The owners have commented that it has never happened before. The tidal Trent is notorious for this. Constant running for hours on end at higher than normal revs.

 

Jen

I think it might be more complicated than this. Steel is a much better conductor of heat than water so the baseplate itself is part of the cooling system, rather than just an unwanted barrier between the internal and external water, and so some heat will run along the baseplate before moving into the canal water. So, even if the baseplate was 50mm thick I suspect it might not be a limitation to the cooling performance. The area consideration for cooling panels likely relates to getting heat IN to the baseplate rather than getting that heat OUT of the baseplate. Suspect the maths would be a bit hard.   Will try some infrared temperature measurements of the steel adjacent to the cooling panel next time I am in a scientific mood.

 

...............Dave

[Jen-in-Wellies]

8 minutes ago, dmr said:

I think it might be more complicated than this. Steel is a much better conductor of heat than water so the baseplate itself is part of the cooling system, rather than just an unwanted barrier between the internal and external water, and so some heat will run along the baseplate before moving into the canal water. So, even if the baseplate was 50mm thick I suspect it might not be a limitation to the cooling performance. The area consideration for cooling panels likely relates to getting heat IN to the baseplate rather than getting that heat OUT of the baseplate. Suspect the maths would be a bit hard.   Will try some infrared temperature measurements of the steel adjacent to the cooling panel next time I am in a scientific mood.

 

...............Dave

Yes, this is all one dimensional. Heat transfer through the thickness, but a reasonable approximation when the thickness is very thin compared with the area as in, for example, a skin tank! In this case, a 6 to 10mm thick plate over an area of 600 by 1800mm. 3 dimensional heat transfer sums are hard and often best done by finite element modelling on a computer. The heat source and sink are both flowing water, so good for transferring heat on each side with large volumes of liquid flowing past, especially on the canal side. More convection than conduction for heat transfer in the coolant and canal water.

 

Jen

Edited August 1, 2018 by Jen-in-Wellies

[dmr]

5 minutes ago, Jen-in-Wellies said:

Yes, this is all one dimensional. Heat transfer through the thickness, but a reasonable approximation when the thickness is very thin compared with the area as in, for example, a skin tank! In this case, a 6 to 10mm thick plate over an area of 600 by 1800mm. 3 dimensional heat transfer sums are hard and often best done by finite element modelling on a computer. The heat source and sink are both flowing water, so good for transferring heat on each side with large volumes of liquid flowing past, especially on the canal side. More convection than conduction for heat transfer in the coolant and canal water.

 

Jen

NO no no, FE modelling was a thing of my past and I am not going to do it again, not ever, still have nightmares about those damn triangles ?.

 

We have a floor mounted panel and it can cool our JD3 (about 30Hp) running flat out. Not a clue how I would ever drain it though.

 

...............Dave

[nicknorman]

2 hours ago, john6767 said:

Don’t all Hudson’s have the skin tank on the base plate, so I guess it works, but possibly more difficult the bleed.

No. Ours has the normal skin tank on the swim. The fuel tank is under the engine, integral with the baseplate however. Maybe Hudsons with mid-engines have skin tank in the baseplate? They have to be pretty thin so than the hot water doesn’t just flow over (under) the upper surface though. Perhaps that and them being the low point in the system, might lead to a risk of sludging up?

[Waterway2go]

If all else fails, try this extreme solution - a cooling "pod" suspended below the counter.     This was fitted to the Yarrow Seagull when she returned to the shipyard that built her for a 25 year refit, with a more powerful engine.   Yarrow Seagull operates on the Forth & Clyde Canal and is one of 10 boats owned by  Seagull Trust Cruises, a charity that carries disabled people on the Forth & Clyde, Union and Caledonian Canals.   Despite my initial misgivings, the cooling pod has not yet hit the bottom, or collected any significant amounts of weed or refuse.

[ditchcrawler]

Biggles who use to post here made one http://boatbuildblog.blogspot.com/2013/08/making-keel-cooling-for-friend.html an interesting read

 

[Tony Brooks]

10 hours ago, Waterway2go said:

If all else fails, try this extreme solution - a cooling "pod" suspended below the counter.     This was fitted to the Yarrow Seagull when she returned to the shipyard that built her for a 25 year refit, with a more powerful engine.   Yarrow Seagull operates on the Forth & Clyde Canal and is one of 10 boats owned by  Seagull Trust Cruises, a charity that carries disabled people on the Forth & Clyde, Union and Caledonian Canals.   Despite my initial misgivings, the cooling pod has not yet hit the bottom, or collected any significant amounts of weed or refuse.

 

 

That's just a keel cooler in a fancy protector. Personally I think the OP could do something similar without the protector but fix it to the outside of the swim. The pipes in the photo in my opinion stand down too far. they only need to be just clear of the hull.

 

[Boater Sam]

Seen a few boats with pipes close up under the uxter plate along the swims, would think that the surface area was too small to be worth the installation.

[BEngo]

I have a run of 4 x 9  ft lengths of 1 inch steel pipe in a zigzag as extra cooling. Was much quicker to fit than building a baffled skin tank.

 Work out the outside area of a foot run of 1 inch pipe, rembering that 1 inch is the ID. You might be surprised.

N

[Tony Brooks]

45 minutes ago, Boater Sam said:

Seen a few boats with pipes close up under the uxter plate along the swims, would think that the surface area was too small to be worth the installation.

The official keel cooler for BMC 1.5s by Newage/Tempest consisted of 4 x 4ft 5/8" or 3/4" OD pipes parallel to each other and set in manifolds. That was for full power sea use so as additional cooling two pipes at 4ft would probably be enough and with a zigzag that is easy to accommodate on the swim.

Edited August 2, 2018 by Tony Brooks

[ditchcrawler]

13 hours ago, ditchcrawler said:

Biggles who use to post here made one http://boatbuildblog.blogspot.com/2013/08/making-keel-cooling-for-friend.html an interesting read