engine cooling expansion tank connection.

[sarawnw]

Hi All,

I want to increase the efficiency of the cooling system and reduce the amount of water vapor from the expansion tank.

Back round info:  I have 12/2 Lister that is cooled with a heat exchanger using the typical thermal siphon setup.  There are 190 F thermostats for each cylinder head, with 3/16 bypass holes to let the air thru.  There is an expansion tank connected to the top of the heat exchanger  similar to every one Else's it seams.  When the thermostats open up, a rush of hot water expands into the tank then gets sucked back into the engine and heat exchanger when the water cools.  There is 1" hose from the engine to the heat exchanger with a rise of 1.5 feet.  The return line from the heat exchanger to the engine is just above level with the same type 1" radiator hose. The best I could do with the room I have and the size of the heat exchanger, system works well. 

Question:
What would be the problem if the expansion tank was connected to the cold engine side of the heat exchanger and have a valve at the engine hot side of the heat exchanger to let the air out?  The expansion tank would still need to be the same height to ensure the system is filled with water.

Has this ever been done before and what where the problems?

Why would you want to do this?  More of the hot water would get pushed through heat exchanger to be cooled by the secondary storage tank water and maybe capture more heat.  There would be less water vapor into the air and  still keep the pressure in the primary side of the cooling system low? 


Any all comments welcome, further explanations gladly given. 


Sara 

[Ronmar]

Can you provide some more info on your setup?  You said the rise is 18" from head to top of heatex.  How far of a horizontal run?  What type and size heatexchanger are you runing?

There should be no "rushing" of coolant into the expansion tank, and no vapor escaping from the tank.  This indicates to me that there is boiling somewhere in the system.  I run a heatex with the expansion tank connected to the elbow feeding into the top of the heatex.  The level in the tank changes slowly with engine temp. The only time I have seen what you describe is when I had a slowdown on the secondary loop and wasn't carrying the heat away from the engine. When this happens, the engine will revert to hopper cooling and the heat leaves the engine in water vapor form.  Boiling transfers a tremendous ammount of heat...  If your system is boiling at some point, it dosn't matter where the expansion tank is connected.  The pressure increase from the boiling expansion will still push fluid into the tank.  The purpose of attaching it at the highest point is so that it will automatically expell air, and in a thermosiphon system that is very important as an air bubble at the top will kill all the flow...

One issue I ran into was getting it to thermosiphon well because the mass in the heatex was only a fraction of that in the cylinder and head of my 6/1.  I had to put the heatex almost directly above the outlet on the head, with about a 30" vertical run from head to top of heatex to get enough flow to cool at full load.  I also put a larger metal pipe on the heatex outlet to contain a larger volume of cooled fluid to aid in the balance of the thermosiphon loop.  My avitar is a scale drawing of my system.  Here is larger scale drawing of my system to show the heatex in relation to the engine...  Any lower in relation to the head and it wouldn't move enough heat reliably at full load...  

[sarawnw]

Hi Romnar

Thank you for the reply.
"Can you provide some more info on your setup?  You said the rise is 18" from head to top of heatex.  How far of a horizontal run?  What type and size heatexchanger are you runing?"

The heat exchanger it self is 20" tall by 4" in diameter with two 3/4 inch fittings coming out of the top and bottom for primary and secondary connections.  With the pipe fitting to adapt to 1" fittings, the total height is just over 24" tall.  The heatex came from a company that made silver reclamation equipment for a no defunct silver halide industry in Rochester area.  The unit was brand new and custom made in Spain. There is a 3" by 18" pipe that was supposed to be for filling and expansion on top of the primary input of the heat exchanger. The filling part works great, the expansion part not so well.  I had to add a car expansion tank from a Kai or Spirit I believe. Amazing the amount water will expand when heated. I should have calculated that.

The rise is closer to 14 inches due the addition of the anti vibration mats to keep the neighbors happy. The horizontal run is almost 4 feet and appears to work pretty well. The heatex input temperature is around 200-220 F, output temperature is around 80 to 90 F. The secondary output is around 190F with an input (guessing, just warm) 60 to 80 F depending on how long the engine has been running.

The rushing appears to come from when the thermostats open and close, which operate independently.  The amount of rushing is dependent on whether the thermostats are in or out of phase in operation.  I don't have the measured period between rushing or the time between min and max rushing operation.  I would think this would change depending in load.  the estimated time between expansion and contraction is 5 minutes with the duty cycle around 40% for heat portion.

After reading my description and your experience, I have a restricted primary side or I need better thermostats?  I would prefer to keep the thermostats since I use a coaxial fuel heating system in the upper hose. (more questions) The outer diameter of the fuel line that runs inside the 1" ID hose is 1/4" in diameter.  I have restricted the primary side to much? 

Just to add more complexity to the mix and to change my analysis a little.  I was amazed by how much smother the engine ran on WVO than on Diesel. I also have the carbon build up that others have mentioned which required a tear down every month to month and a half to remove the carbon when running 24/7. I also had more condensation in the exhaust when running WVO than Diesel.  Get ready to cringe,  I adjusted the injection timing on both cylinders from the 18 degrees traditional setting to 22 degrees in 1 degree steps to see if I could get the engine to perform as well as on diesel.  Meaning, have same punch on each power stroke as when using diesel fuel. As it turns out, at 23 degrees just knocks like crazy and 22 degrees runs like using diesel but using WVO. 
There are some problems running at 22 degrees injection timing, starting on Diesel causes the engine to knock.  I start on diesel just to get it running on a cold start and quickly change over to WVO.  The other problem is, I don't create as much heat out of the engine to heat my house which brought me to the question of moving the expansion tank connection placement.

I hear more questions.  The good side of 22degrees injection timing is increased power, 1/3 less fuel usage for same load, much less condensation ( have not figured why out yet) and the holey grail, tear downs due to carbon are 3 months instead of a month.  Some of the carbon reduction could be due to the modified exhaust the uses venturii effect to suck the air out of the non firing cylinder that got changed at the same time. (I know, only change one thing at a time.) Oh, the other good/bad effect is I have to hurry along with maintenance so I don't have to start on diesel.

Even with the reduced heat, there still is water expansion as before and the water temperatures are about the same as before, expansion cycle times are longer come to think of it.  Will measure expansion cycle period when I put the engine back together.  The amount of heat that I can pull out of the secondary cooling 55 gallon drum is about half based on how often the circulation pump needs to turn on.  FYI, The circulation pump pulls hot water out of the drum when top 3/4 of the drum gets to 180F and dumps the heat into the water tank(another heat exchanger) and furnace radiator coil. A simple push system that sends heat whether you need it or not. The furnace heat get connected to the pool pump heat exchanger in the summer.  It's amazing what you can do with garden hoses.  I'm glad my significant other doesn't venture into the basement.

Now that you the full story,  what do you think?

Sara 

[Ronmar]

I think you have too much horizontal component in your thermosiphon system and it is boiling due to inadequate flow.  That 220F at the input to the heatex is one clue, the 80-90F at the return to the engine is the second Those numbers are the highest I ever see on mine, and the thermostat is nearly fully open at that point on mine.  I think your thermostats are not cycling, but open all the time  as soon as the engine warms.  Thermostat cycling is caused by too much available inlet cool water, and too much flow. They are also accompanied by temperature swings below the thermostat set point.  14" high to 48" long equates to a 30 degree slope up to the heatex and  with the height of the heatex, the return to the cylinders is even more horizontal.  I think the only reason it is working at all is that your heatex has a high volume, so the weight of coolant between hot and cold sides is closer to being equal.

You probably have 3-4 gallons of coolant, perhaps a little more?  If you are seeing large swings in overflow height, I doubt it is from expansion.  First off the coolant is already warm.  With a little over a gallon of coolant from cold to full op temp I see about a 1/2" rise in my little plastic exp tank.  Maybe 1/2 cup of volume change...  IF yours is already warm, where is the expansion comming from?  Thermosiphon is a very low energy pumping method. It is powered by gravity and horizontal runs are it's enemy. It won't have enough energy to force fluid up into the tank.  

Boiling on the other hand can transfer a tremendous ammount of heat in a short time.  What I think is happening to you is that the coolant boils in the cylinders.  This boiling and bubbles pass up into the pipe and mix with coolant and the bubbles condense back in with the cooler fluid.  This expansion causes the increases and the contraction causes it to decrease.  The boiling also pulls a little additional fresh coolant into the cylinder from the bottom which kills the boil.  This coolant with it's slow flow, slowly warms untill it again boils and causes the cylinder to gulp in fresh coolant...

At those temps, you don't need a restriction, you need an increase in flow.  I would suggest you keep the same vertical relationshp, but move the heatex right next to the engine and loose as much of the horizontal component as possible...  Is your secondary loop pumped?  

[Tom Reed]

Wow, you've got a lot of stuff going on there. Ronmar may be on the right track, however one conflicting piece of info is that there are slugs of hot water happening. To me this is not indicative of a lack of flow, but to much. I believe a lack of flow would produce a continuous, but slow stream of really hot water. Do you have an IR thermometer you could monitor the outlet temp with?

Keep us informed, it will be interesting to see what you find. Also could you post a pic of your install and coolant line routing? Sometimes we can see things in a pic that haven't been said.

[XYZER]

A poor thermosiphon will cause a lack of flow until steam is formed. Then it will really flow

[mobile_bob]

are you using a pump on the secondary side of the heatex?

the reason i ask is this

if you are using one, i would suggest getting rid of both t/stats
just let the coolant circulate unimpeded on the primary side of the exchanger

and control the the amount of heat retained in the engine by use of a temp switch to control the secondary side water pump.

in this manner the primary side could be setup to run up to the switch limit temp, the pump would come on and heat removed until the switch opens again. meanwhile the flow in the primary side just keeps right on doing its thing without restriction or any throttling do to opening and closing of a t/stat (or in your case two t/stats)

just a thought.

bob g

[Ronmar]

I believe a lack of flow would produce a continuous, but slow stream of really hot water. Do you have an IR thermometer you could monitor the outlet temp with?

XYZER is right on.  A lack of flow will produce a stream of really hot water, untill the boiling point is reached.  Then freight trains of bubbles are formed that pull a bunch of coolant along with them.  Ever see a salvage diver using a surface suction dredge?  They inject air at the bottom of the hose, and it carries water and the sand/silt up and away from the work site...

Temps above 210F at the heatex inlet with 190F thermostats tell me low flow...

Pics of the system would be great...

[sarawnw]

Hi All,
I apologize for not replying sooner, work and family keep a variety of things to do in life that can not be denied.
Romnar,
I find your analysis very revealing and ripe with opportunity for optimizations. 
I wish I could move the heat exchanger closer and am thinking of ways to do so.  Due to space , building code compliance, neighbor relations and their prolific expectations, and environmental constraints, doing so and maintaining the height for thermal syphoning of both the primary and secondary side of the heat exchanger is a challenge.
I do have a taco pump which can pull the water from the top of the secondary top side heat exchanger and storage tank which circulates the water to the water heater and furnace heat exchangers. The return line of the taco circulation pump is at the bottom of the storage tank.  The distance between the heat exchanger and storage tank is 4 inches with a height difference of 4 inches. Thermal syphoning seems to works well here based on the fact that I can run the engine 5 or 6 hours before the bottom of the cooling storage tank reaches 120 F   and feel I have to shut down or remove heat from the storage tank. 
The primary side coolant capacity is just less than 4 gallons based on when I initially fill the system after tear downs.  If I increase the volume of the primary side cooling system, say another gallon, will the hot water be pushed enough through the engine, effectively increasing flow,  to keep the water from boiling?  Other than extra stress in the radiator hose, what other problems can I expect when allowing the water to boil in the engine heads? Pressure is reduced almost immediately or shortly thereafter due the expansion tank.  The head temperature will be around 212 F if the water is changing between gas and liquid states in a low pressure environment provided there is enough cool water available to maintain system equilibrium. Using the boiling water to pump the water into the heat exchanger seems like a condition to take advantage of if the pressure can be controlled.  Could there be problems with head gasket, injection pump, lubrication of valve train, etc. using boiling water in the head?   
What should the slope be for thermal syphoning to work correctly? Is there  a rule of thumb for 1 inch pipe or hose? 
From the spit on the finger sizzle test, the only part of the head that sizzles is the exhaust port and the injection pipe just after the heater and before going into the injector, the IR temperature readings confirm this. Can I assume the boiling is most likely contained to with in the head which means the quantity of water boiling is relatively small compared to the four gallons of coolant?   

I am going to remove the thermostats for the next run before the maintenance tear down.  Will be looking around for other ways to shorten the horizontal distance.

Tom,
I believe Romnar is right about what is happening and glad we have him as a resource. The volume of expansion that I am seeing doesn't make sense compared to thermal expansion of water Romnar experiences.  The casting that connects the two head water inlets or outlets by function on the 12/2, forms a lip of ½ inch or so between the water in the head and the top of the common water fitting on the casting. This ½ inch pocket of water is higher than the water outlet and is not allowed to escape.  I don't know if this is common to all 12/2 engines but might explain a small volume of water that could boil given low flow cooling system as my system provides.  The IR temperature readings at output of engine are 200 to 220F with 205 to 210 being typical readings.  I'm looking for picture since I tore it down for maintenance.  Whenever I do, brain starts racing of how to improve the system to a achieve an autonomous power plant.
Bob,
The circulation pump effectively does work as you describe but with a delay of the top of the cooling tower providing a buffer to incorporate some hysteresis.  An interesting way to look at how the cooling system works and how to maintain system temperature.  The thermostats would really only allow the engine to warm faster. If the engine is creating boiling water anyway due to lack of flow,  thermostats are redundant and actually hinder the pumping action of the water/gas transitions cycles.
Thanks again for all comments and valuable insights.

Sara

[Ronmar]

I would also reccomend pulling the thermostats as a first step.  At those temps, they are wide open anyway.  Removing them should improve flow a little.  Since you are cyclically boiling and not at a steady boil, that may be enough to keep you just below that boiling point.  As long as it is receiving enough replacement coolant to control the boil which appears to be the case, I don't see a serious problem other than the temperature rapidly cycling.  It is better to maintain a steady temperature from a wear and metal fatigue viewpoint. Since the system is vented thru the expansion tank, there should not be any pressure issues, just the rapid changes in expansion tank level.

Anything you can do to smooth out the pathways between the engine and heatex. The most direct path is desired.  High spots and sags should be avoided.  The highest point in the system shuld be at the heatex inlet. 

I also don't see a problem of temps in the low 200's from a combustion standpoint.  IMO, the engine will run more efficient and burn cleaner at these temps than say 170-180F.

Another option would be to add a circulation pump to the primary loop.  This would increase the flow and put the thermostats back into the equation and they would maintain temp untill the return coolant to the engine increased to the point where it could no longer cool effectively due to secondary heat saturation...

DId I read correctly that your secondary loop pump is on a thermal switch and cycles?

Some pics would really help us to help you...