Coolant loop flow measuring

[quinnf]

Jens,

I think you'll find that the total amount of heat you can recover from the coolant and exhaust will be more than you expect.  But I wouldn't obscess too much looking for a nice clean 1/3 distribution.


To continue the earlier discussion, (I don't mean to argue; I think it's just an interesting point), you said:  "In both the engine and the exhaust heatex the coolant removes the heat at the same rate it is generated.

Not quite.  In both cases, the coolant removes heat at the same rate is is conducted, not generated.  And the total heat of combustion, less work done, is divided at least three ways, not two, as the 1/3 rule would have it.  

Last point first:  The 1/3 rule of thumb assumes heat is evenly divided between coolant and exhaust gas and ignores heat lost to conduction and radiation (and vibration/noise, material flexure, etc).  Recall that these engines burn very little fuel for their mass.  There is a lot of metal to conduct heat away from the combustion chamber.  

The 1/3 rule is a really rough rule of thumb.  It should probably be stated:

Some heat is translated into work, (18-20% best case)
Some heat is conducted into the coolant.  
Some heat passes into the exhaust stream, and,
Some is conducted away into the metal of the engine and is lost to radiation and convection.  (I've read ~10% several sources)


My earlier statement about casting thickness didn't convince you.  
Imagine two identical engines:  One with thick metal castings, and another one with thin castings.  

Fourier's Law* holds that the quantity of heat transmitted through conduction is inversely proportional to the thickness of the material through which the heat is conducted.  So of the two engines, the one with the thicker castings will conduct less heat per unit time into the coolant.  That means more heat will remain in the combustion chamber, resulting in higher temperature and pressure than would be the case in the engine with the thinner castings.  There would be no continuous heat buildup leading to a "meltdown" because after each combustion event, the exhaust valve opens, and the combustion products pass into the exhaust stream, where more heat will be available to transfer into the coolant.  

On the other hand, the engine with the thinner castings would transmit a greater proportion of the available heat of combustion to the coolant, leaving less heat remaining to pass into the exhaust stream.  

So, given the relatively heavy construction of the 6/1 type engine, I'm not surprised that more heat than was expected can be found in the exhaust.  


*Q=kAdT/s

Q   Heat transfer per unit time
A   Heat transfer area
dT Temperature differential
k   Thermal conductivity
s   Material thickness



Merry Christmas, Everyone!

Quinn

[Ronmar]

Jens
  You mind publishing your graphs?  Of especial interest to me would be the input and output temps of he EX heatex and the engine...  Also what is the size difference between your exhaust heat exchanger surface area and your engine coolant surface area?  That could explain some of the differences you are seeing. 

You are saying that the Exhaust heat exchanger is collecting twice the BTU as the engine coolant, but you have stated several times that you don't have the ability to measure the BTU output? Are you basing this assumption on strictly the thermal rise across the devices?   If you go ahead and measure the actual BTU output, you wll probably find that the EX heatex is not collecting twice as many BTU as the engine.  But in reality, the engine coolant is only collecting 1/2 the BTU that the exhaust heatex is...

Heat transfer is about thermal coefficient of the barrier material, surface area, flow, contact time, turbulence, and probably most important to this, the temperature difference from one side of the barrier to the other...  In your case, The coolest fluid arrives at the exhaust heat exchanger and is exposed to the highest heat source.  The exhaust gas also passes into the heat exchanger and expands and slows down, so it is in contact with the heatex surface area longer than in the combustion chamber.  The wall thickness as Quinn mentioned is also thinner.  All this and the high temp difference leads to a very high heat transfer rate for your given fluid flow, so no wonder you see a much larger temperature gain.  You then take this heated coolant out of the heatex and pump it thru the overall lower temp of the engine cooling passages.  Lower Delta = lower/slower heat transfer. 

If you swapped the flow so it entered the engine first, then went to the EX heatex, You would probably see a reversal in temperature gains for a given flow.  If you split the system and fed both "heaters"   with the same temperature fluid, you would probably see comparable temperature rises for a given flow. 
I think you are comparing apples to oranges...

[Geno]

If you swapped the flow so it entered the engine first, then went to the EX heatex, You would probably see a reversal in temperature gains for a given flow.  If you split the system and fed both "heaters"   with the same temperature fluid, you would probably see comparable temperature rises for a given flow.  
I think you are comparing apples to oranges...

That's what I've been doing for a 1000+ hours and it does exactly as Ron says. I see nearly equal temp gains in both of my heat exchangers. The coolant temp return to the engine is steady at ~140°F after running all day. I do have a thermostat regulated fan on the coolant return to the engine but it rarely turns on.
Ron and others helped me work the bugs out of my design and it made a big difference in all aspects. What has being advised here works, is the most efficient, time tested and easiest to implement design.



Thanks, Geno

[Ronmar]

If you maximize your heat transfer to fluid, you should actually have to run the engine less(more BTU transfered in a shorter time)... Since your fuel consumption is fixed with load and time, the faster the transfer, the better.   What storage temperture are you looking for?  Your engine is capable of providing up to 220F before it hits it's shutdown point, and probably pretty close to 195F over the long term average.  You mentioned 3 tanks, 80 gallons each, for storage and some of that is used for domestic hot water, so surely you do not want over 120F-130F for safety reasons, or do you have a TRV on the domestic outlet?

Well 195F to 120F is 75F of delta at the worst case.  80F in the tanks would be 115F of delta.  The transfer heat exchanger should be sized to deal with these working deltas to transfer the heat in the time allowed.  Based on your storage sizes, we can figure roughly what you should see for run times based on a known energy input.

Say 5KW electrical load @ .125 gallon per KW/HR.  Diesel is around 140,000 BTU per gallon, so a 1 hour run at 5KW will consume roughly 87,000 BTU in fuel.  2/3 of that is around 57,750 BTU that you should be reclaiming.  You have 1920 pounds of water in storage, so that reclaimed heat should raise the temp of that water roughly 30F(1BTU raises 1LB, 1 degree F).  How quickly it does it depends on the size of the heatexchanger and the delta across it.  Got low delta?  Need a bigger heatex to compensate for that design parameter... 

I think you mentioned that the 3 tanks were plummed in series?  This also complicates things somewhat as the tanks tend to separate vertically by heat.  So you are pushing hot water into one which is pushing it's, I assume coldest water into the next tank and so on, so you will have cooler pockets in each tank, till ALL the water gets circulated.  Ideally you would feed all 3 tanks in parallel and feed the hot water into the top and draw the coolest water out of the bottom to provide the coolest water to the heat source/exchanger to maintain the highest delta. Parallel tanks however provide their own set of flow issues...

[oliver90owner]

Don't forget there is considerable energy imparted as kinetic energy to the gas flow.  Turbos recover a fair part of this for boosting engine outputs.  Yes, knock on effect, extra boost equals more exhaust velocity eqals more recoverable energy, etc.  But it is there and is normally lost on a 6/1 engine as pumping losses.

Only added this as an extra way that the energy of the fuel is transferred out of the engine, lesser losses would be as sound energy.  So the total energy balance of the system is complex to say the least.  Rule of 'one thirds' is a pretty approximate one. Just like the rule of 2HP per kW.

Regards, RAB

[Ronmar]

Jens, Nice Graphs!  Are those output from your data collection software?  What are you using to collect the data?
  I would say you do indeed have some sensor error/calibration issues.  If you have a 96C/205F thermostat in the engine  and the highest temp seen on the engine output graph is 90C...  That is a big difference.  I have no difficulty reading the coolant temp at the engine outlet fitting with an IR temp gun on my engine, and it is pretty much within a degree or two of the thermostat temp, so the heat is available there for measurement.    Have you compared the sample points with an IR shot of the same area?  The sawtoothing on the engine graph is indicative to me of the thermostat cycling, so the temp at that point is probably around 96C. That is maybe a 15C error from what is seen on the graphs? Since the thermostat controls overall flow, you can also see the fluxuations on the heatex graphs, so you are capturing those changes nicely.  Right up to the point, that the engine inlet temp rose high enough that the thermostat went full open, and the temps started to climb away smoothly from there.

The fact that you are able to see the thermostat cycles tells me that you are getting pretty good thermal transfer to the sensor.  Is there a calibration factor adjustment in your data collection device/software to calibrate the sensors output to the collection device input?

What was the data sample rate timing on these graphs?

Measurment consistency can,  as you see now, really make a big difference in your results.  By that last graph, I would say that the engine is delivering approximately a 50% greater heat rise than the EX heatex is.      

[Ronmar]

Well if the temp IS accurate, you are missing out on another 15C of possible delta on the generator output to the secondary heatex.  But the temp being accurate dosnt make sense after seeing the graphs.  I have found thermostats to be pretty consistent devices. Take the temp gauge in your vehicle as an example.  The engine output graph shows a steady temp increase at first(water pushed thru thermostat vent hole?).  Then it starts a series of rapid cycles and the temp graph mostly levels off.  That indicates to me that the thermostat has reached it's temperature and is cycling to maintain it.  Now if it was a 80C thermostat, I would say the temp readings would back that up.  But a 90C thermostat isn't even open at 80C...

[Ronmar]

Yes, you saw 96C on that last graph, but not untill the very end of the run. The thermal cycling that appears to me to be thermostat modulating the temperature stopped long before that, back in the low 80C range...

A couple of other observations I was thinking about.  The short cycling on the engine output graph might indicate a little too much flow on the primary loop.  A little less flow might allow the thermostat to open to a point and remain there. A low pressure gauge on the pump output might also show these thermostat cycles as slight pressure spikes every time the thermostat closes. 

Boiling in the secondary loop with 100C input indicates perhaps not quite enough flow in the secondary loop.  With 100C input, you should have something less than that on the output.  The secondary loop enters at the storage tank temp.  But with enough flow, the secondary loop heatex outlet should still be less than the primary loop inlet.  No temp difference/little delta, no heat transfer...  With the storage temp still well below 100C, you are not moving enough fluid to keep the heatex from boiling.  This low flow is costing you potential delta/transfer rate near the end of your run..

[Ronmar]

Yep, the losses from a high storage temp are a problem.  A larger tank at a lower temp will store the same ammount of energy, but have less loss for a given insulation thickness.  It is also hard to get that last heat into the tank when the tank storage temp is approaching that of the heat source:)

[Ronmar]

Fair enough ... but remember that I don't know the exact coolant temperature at which the secondary on the heat exchanger starts boiling.

Jens

Well if there is any secondary flow at all, the coolant temp reaching the heat exchanger must be somewhere over 100C.

Do you have graphs of the secondary loop inlet and outlet temps?

[Crumpite]

Folks,

Just a quick heads-up on measuring flowing liquid temperatures:

Even with the most modern (as in $$$) equipment available, chemical engineers need to resort to a lot of hand waving when trying to get an accurate energy balance on small scale systems (like ours)

Technically, the best way to get a temperature of a flowing liquid it to use a very small temperature sensor inserted into the center of the flow.
Usually the leads up to the sensor are very thin to minimize heat loss through the leads.
Then you apply a correction factor that includes the thermodynamic data on the liquid flowing, it's flow speed, distance of the sensor from the edge of the pipe, what type of pipe it is, phase of the moon and whether you got a haircut that day.

So don't feel bad if you can't make hide nor hair of what's exactly going on inside your CHP unit !
I've seen some engineers fresh from collage getting the edges of their diploma's dogeared from chewing on them due to a 'simple' energy balance on a small system.

Don't get discouraged, it *will* make sense after you do enough playing around, (at least it usually does...)
Daryl