Keel as Heat Exchanger

[sea_casper]

Can anyone explain why keels with their huge thermal mass and surface area are not utilized as the heat exchanger for engines, refrigeration,air conditioning etc.?
For all applications a closed system is more reliable and requires less maintenance.It would eliminate raw water pump, engine mounted exchangers, raw water filters etc.

Even an add on system attached to the side of metal keels would not seem too difficult an engineering exercise.

 

[RAD]

First thing that comes to mind is drag.....a tube running fore and aft right behind the keel might work

 

[Merlin Clark]

There are systems available called "keel coolers" for engines and refrigeration (Frigoboat) and probably AC systems that mount to the hull.
They've been around for a long time and work great. No idea why they aren't more popular except maybe the industry just makes too much money off of replacing or repair of the more complicated and expensive systems.
http://www.waltergear.com/kc.htm

 

[RichH]

The problem is the 'approach temperature' of all heat exchangers ... typically 15° (the terminal temperature difference between inlet and outlet of the exchanger) X something called log mean temperature difference X surface area X the coefficient of heat transfer. This would require a very complex metal casting of the (lead) keel - (complex cooling passages) and a keel that 'shouldnt' be painted in those areas (or degradation of heat transfer).
Its just too damn easy and economical to 'optimize' all the 'variables' of heat transfer in a simple and relatively small device that simply 'attaches' to the engine; the 'tube bundle' of these small Ht. exchangers are already close to the surface area of most keels.

What you propose is do-able but the expense to do it 'right' would be quite costly and would be quite hard to 'maintain' at a constant operating regime - lead has a quite relatively poor value of heat transfer in comparison to thin walled stainless or (admiralty) bronze tubes that one finds inside a simple heat exchanger, and one can pack a relatively large 'surface area' (tube surface area) inside a ht. exchanger. Also think about the bio-fouling that constantly forms on keels, etc.

 

[sea_casper]

Thermal Conductivity

The problem is the 'approach temperature' of all heat exchangers ... typically 15° (the terminal temperature difference between inlet and outlet of the exchanger) X something called log mean temperature difference X surface area X the coefficient of heat transfer. This would require a very complex metal casting of the (lead) keel - (complex cooling passages) and a keel that 'shouldnt' be painted in those areas (or degradation of heat transfer).
Its just too damn easy and economical to 'optimize' all the 'variables' of heat transfer in a simple and relatively small device that simply 'attaches' to the engine; the 'tube bundle' of these small Ht. exchangers are already close to the surface area of most keels.

What you propose is do-able but the expense to do it 'right' would be quite costly and would be quite hard to 'maintain' at a constant operating regime - lead has a quite relatively poor value of heat transfer in comparison to thin walled stainless or (admiralty) bronze tubes that one finds inside a simple heat exchanger, and one can pack a relatively large 'surface area' (tube surface area) inside a ht. exchanger. Also think about the bio-fouling that constantly forms on keels, etc.

Lead has a Thermal Conductivity Watt / (m-C) of 35, Cast Iron is 55, Stainless is 16, Bronze is 26, lead is not that poor.

Embedding a Copper coil in the lead casting would not be difficult, certainly not intricate, although I would prefer a removable flat section attached to one side, as it would be easier to repair if needed.

Embedding the cooling within the keel thermal mass would add no additional drag. The lower conductivity of lead is more than compensated by the thermal mass and large wetted area, which transfers heat to the surrounding water.

Standard anti-fouling paint would not have an adverse affect on cooling because any reduced thermal efficiency is overcome by the sheer size of the keel. You trade efficiency using size, for an engine mounted heat exchanger you trade size using efficiency.

 

[RichH]

Well if you insist, just perform the simple heat balance calculations applied to 'heat exchangers' and see what your "Uf coefficient" turns out to be when using THICK walled lead passageways. putting a thinwall construction in lead on/near the 'surface' of the keel will severely reduce the keels 'structural moment of inertia' (I^3) ... not including 'stress risers'.

Suggest you can start with the classic: "Heat transfer" by Kerns as a starting point to do your calcs.

k = Btu/(hr) X sq. ft. X °F/ft.

k
Brass/bronze = 60
Stainless = 26
Cast iron = 30
Lead = 19

60/19 = 3.2 !!!!! and you will need to severely increase the 'thickness' based on normal pressure/keel stress when using 'lead'.

Tensile kpsi
Bronze 40-130
Stainless 90
Cast iron 18-60
Lead .6 - 2 !!!!!!!

Est. (SWAG) calc. ~ SA re. Thickness for 'efficiency' =
100/.6 X 60/19 = 526 Bronze in this application will require 526 times 'Less' ('surface area X WALL thickness') than when using lead.

 

[kloudie1]

Some numbers to crunch on if one is so inclined.. The Yanmar 3GMF heat exchanger has a heat transfer area of 1.75 square feet (from their specifications).. To get that same area in 1" tubing, ya need only a bit over 5 feet of coil.. so if the heat transfer is very bad, a pre-formed pressure tested coil that was, say 15 feet long, embedded in a lead or cast iron keel would probably do the trick.. Like Rich says, not an easy thing to do because the coil would have to be braced very carefully in the keel mold to withstand the pour and flotation forces..
Keel coolers are common on steel hulled boats where they box in an interior space on the hull and that becomes the heat exchanger.

 

[sea_casper]

Use Flat Plate Analysis

Forget tubes, consider a flat plate 3.9" thick or 1/10 m and a contact area of a square meter.
Q = heat (watts)
k = thermal conductivity (watt/m °C) Lead=35
Ac = contact area (m2)
T = temperature difference (C) assume 27 C
t = material thickness or length that the heat has to travel (m) assume .1 m

for Lead .1 m thick
Q= kAcT/t = 350 watts per square meter per degree C

If we assume 27 degrees for a temp difference we get 9450 watts per square meter.

If we assume a two flat plate sandwich 18900 watts/m2

A Yanmar 3YM30 engine is rated at 20kw, and only a fraction of that is turned into heat.

So a flat relatively thin void in the center of the keel of 1 square meter is more than enough.

 

[RichH]

The functional differences between a flat plate or tube bundle external and a boat with a 'steady state' ht. exchange system is ... one has ensure the mass flow rate of water though the 'piping' vs. the boat speed at 'zero' or when no water is moving across the plate or tube bundle. The worst case is when the keel is not moving through the water and the 'heat exchange surface' loses its definite or designed "log mean temperature difference' LMTD (and ignoring the small amount of 'convection' of outside cooling water passing by the 'plate', etc.).

The 'beauty' of a traditional Hx system is that its designed to be a steady state exchanger (designed for max. expected heat transfer load) and which doesnt depend to the boats movement through the water.
Simply stated, a keel cooler has to or should be designed with the thought that boat isnt moving, the non moving water will create a decrease in terminal difference ...terminal differences is what makes a heat exchanger 'work' with any efficiency.

I would disagree with a small flat 'void' in the center of the keel due to the THICKNESS of the material and the outside water .... would promote very poor, very inefficient coefficient of transfer ... simply because of the thickness.

Im beating a dead horse, so I'll stop now even that I spent meny years engineering shell and tube, etc. heat exchangers in my 'early days'.

:bang:

 

[kloudie1]

Yup, Casper.. that probably would work too.. I am partial to tubes because that is what I worked with lots over the years.. Much better pressure containment than flat plate but at 6-10 psi , probably not a biggie.. easier to keep from leaking too. but the plate idea would work.if the details are addressed..FYI, . The number that Yanmar has for the little heat exchanger is 0.0072 square meters per horsepower.

 

[sea_casper]

Yanmar undersized theirs.

Yup, Casper.. that probably would work too.. I am partial to tubes because that is what I worked with lots over the years.. Much better pressure containment than flat plate but at 6-10 psi , probably not a biggie.. easier to keep from leaking too. but the plate idea would work.if the details are addressed..FYI, . The number that Yanmar has for the little heat exchanger is 0.0072 square meters per horsepower.

I envision a U-shaped void so inlet and outlet would be on top. Pressure containment and flow requirements would be easier, and less susceptible to clogging than a thin tube exchanger.

The Yanmar heat exchanger is notorious for overheating on my engine the 3ym30, plus I am tired of cleaning the raw water strainer every week, we have eel grass here that tends to get sucked in.

If a plastic sandwich bag ever gets sucked in, you overheat fairly fast.

Also the engine has limited adjustment on the raw water pump belt, after readjusting a new belt after the first week, I'm at the end of its adjust capability.

I believe it is a better systems approach, because of the sheer size available, you don't need high efficiency, only brute force simplicity.

I think RichH would at least concede it would work fine for refrigeration and air conditioning.

 

[Ross]

Re: Yanmar undersized theirs.

External heat exchangers are common on work boats as a set of tubes mounted tight to the hull and circulating engine coolant. One inch tubes about twelve feet long on each side of the bilge is the usual arrangement.

 

[sea_casper]

Addons are Addons

External heat exchangers are common on work boats as a set of tubes mounted tight to the hull and circulating engine coolant. One inch tubes about twelve feet long on each side of the bilge is the usual arrangement.

I know external devices can be used, my point is, a sailboats keel is an asset that can be utilized for a cooling function, without adding drag.

 

[alexco38]

It's a "business" decision. For sailboat builder, you have to partition your suppliers as you buy a few engines for your product. You'll let the engine vendor to doe their design and not have to couple the keel builder and the engine builder in your project plan.

However, if are building lots of the same boat, you certainly do that. Next time you go on a cruise ship, look under the lifeboat. They use an external tube for engine cooling. Because the lifeboat maker make hundreds of the same boat. They can design the engine with the cooling system together. They could save a few dollars as there is no need for a HX.

 

[Ross]

I know external devices can be used, my point is, a sailboats keel is an asset that can be utilized for a cooling function, without adding drag.

There are many "improvements" that can be made to a boat if cost is no objection. Drag on a high horse power work boat is of little consequence. The drag induced by pipes that run parallel to the direction of travel will be small.

 

[kenn]

I know external devices can be used, my point is, a sailboats keel is an asset that can be utilized for a cooling function, without adding drag.

I agree that it is theoretically possible ... but it would raise the cost and hassle of casting a keel significantly, and I suspect it would be economically impractical to accomodate in a mass-produced boat. Two more holes to align and drill into the bottom, two more openings to keep water-tight, and what do you do when something in the keel circuit or the fitting goes bad?

 

[sea_casper]

Simplicity wins

I agree that it is theoretically possible ... but it would raise the cost and hassle of casting a keel significantly, and I suspect it would be economically impractical to accomodate in a mass-produced boat. Two more holes to align and drill into the bottom, two more openings to keep water-tight, and what do you do when something in the keel circuit or the fitting goes bad?

I'm not an expert in keel casting technology, but it's not rocket science, I doubt casting a void, with two fittings stretches it. The simpler the approach the more reliable it is. There are many ways to skin a cat, once the decision is made to incorporate the cooling function in the keel, it will be cost effective, especially if it increases engine reliability The raw water cooling "system" is the weak point in today's marine diesel engine technology. Ever price a Yanmar heat exchanger?

 

[Benny17441]

The premise is sound but it is not practical. I'm just thinking about a sailor half way around the world that has invested thousands of dollars to get to location where he is at and is glad that UPS can deliver a replacement heat exchanger in a matter of days at a relative inconsequential cost. What is that man to do if he had a built in keel heat exchanger that has failed and needed replacement? Let's not say it could not happen.

 

[sea_casper]

Any keel failure

The premise is sound but it is not practical. I'm just thinking about a sailor half way around the world that has invested thousands of dollars to get to location where he is at and is glad that UPS can deliver a replacement heat exchanger in a matter of days at a relative inconsequential cost. What is that man to do if he had a built in keel heat exchanger that has failed and needed replacement? Let's not say it could not happen.

He is in the same position he would be in if a keel bolt fails, or the keel cracks. What type of failure do you foresee? What kind of failure would affect a void in a casting?

 

[wonkodsane]

Loading induced stress cracks between void sections?

Could allow communication between those spaces, dropping cooling efficiency or worse, damaged material creating a blockage?

I like the idea in general, though.