Coolant loop flow measuring

[Jens]

Is there a way, as far as anyone knows, of measuring coolant flow without actually opening the circuit ?

Here is where I am coming from ...... the coolant loop for the engine goes from the engine hot end to a 205 degree thermostat, into the house, through a heat exchanger that cools the coolant, through the exhaust heat exchanger and then into the cold end of the engine.
I have been looking at temperatures in the system and I have come up with an oddity. The coolant, as measured at the input of the cooling heat exchanger in the house, is at about 80 C when the engine temperature exceeds the setpoint for overtemp shut down. I am expecting the thermostat to be wide open and I expect to see a temperature close to 100C at the cooling heat exchanger (instead of the 80C I am seeing). The engine shuts down at 220F so in theory, even if one assumes a small loss of temperature from the uninsulated pex run, 210F/100C or thereabouts seems reasonable.
I measured the output as the coolant exits the engine enclosure (about 2 feet past the thermostat) and I see about 85C there at the moment.
I am having a bit of an issue trying to think what is causing this temperature. I can see the temperature gradually heating up in the coolant run but anything coming from the engine should be (according to my theory anyway) be close to the rated temperature of the thermostat in the engine.
There is one possible issue in that the coolant hose past the thermostat takes a big dip before exiting through the top of the engine enclosure. Could it be that all the heat is trapped there ? I never thought this would be a problem since this is a pumped circuit and not a thermo-cycling type setup. In addition to the droop in the engine enclosure, the coolant hose makes another large dip a bit further and I can't eliminate that at the moment.

Anyhow, I am trying to figure out why I would see 220F just past the thermostat and only 85C only 2 feet further along the coolant path.
The only possible explanation I can think off is that the flow through the system is very slow. I am using a Taco 007 pump which is located between the exhaust heat exchanger and the cold side of the engine but I have no idea how fast the coolant  is actually circulating. I will be installing some more sensors to check temperatures at various points which might help explain things a bit but flow is an unknown. I suppose I could install a Taco 012 which has a much higher flow rate and head but I would rather not just guess at this issue.

I would welcome any thoughts about the possible causes of the observed temperatures and/or thoughts about how to track down if there may not be enough circulation going on.

I do have a spare Taco 007 so that I could put two in series but I would hate to just try things because I use straight antifreeze and I always end up with spillage.

Jens

[oliver90owner]

Taco XXX - presumably this is an electric central heating pump?  If so there will be published data on head/flow/power requirement/etc.

Regards, RAB

[oliver90owner]

Oh, and your comment about antifreeze.  I do hope that is propylene glycol and not ethylene glycol.  The former is a food grade chemical, the latter is very poisonous.  Any hint of a leak into water, that might possibly be used for drinking, is a risk not worth taking.

Install a drain-down point, then there will be no need for anything but the smallest spillage.

Regard, RAB

[Geno]

I assume your using an IR thermometer to read temps. If so are the materials your shooting the same? Same color? Same texture?

[rcavictim]

For a simple flow indicator try to find one of those sight glasses from a decommissioned gasoline dispensing pump.  It has the petalled butterfly thing inside that spins around.

[cognos]

I'm pretty sure you all know this...

In order to use most flow spec charts, you need a discharge pressure at the outlet of the pump, and a suction pressure at the inlet. Subtract the inlet pressure from the outlet pressure, this will give you a poormans' net positive suction head pressure (NPSH). Multiply that pressure by 2 (roughly accurate for water) to give you head in feet, and use that number on the flow spec chart to get a more real-time flow measurement.

This will also give you a fair(er) representation of the restriction to flow in your system.

I know it's not really realistic to do in a DIY system - but all exchangers should have a pressure gauge on the inlet and outlet, or the pressure should be measured every now and then to see if anything's changed that would indicate fouling or plugging or a pump going bad...

I would find it difficult to believe you are losing 15­°C in 2 feet of hose... I would supect an air trap at the measurement point - you're measuring the temperature of the air instead of the liquid, or a faulty measuring element...

(I know you must have a good reason, but I'm still trying to figure out why you bring your return flow from the domestic ex back through the exhaust heat exchanger just before reinjecting it back into the engine... I would have thought that the exhaust heat ex should be just after the hot water comes out of the head, just before the exchanger that you use to pull out the heat for domestic use. In your system, you are preheating engine coolant...  Is the coolant so cooled by the domestic ex that you are worried about thermal shock to the engine? Or - would passing it through the engine, then the exhaust ex raise the coolant temp above the operational limit of the piping/equipment/boiling point of the coolant?)

[cognos]

Ah! Okay. I understand why you did it that way.

But - I think ther might be a flaw in your logic. It doesn't matter what temp the inlet is  - hot or cold makes no difference as to how much heat you will pick up - as long as there is no state change...

What I mean is, if you heat water from 40°C to 80°C, it gains as much heat as if you heated it from 50°C to 90°C - the net change is 40°C in both cases. As long as the flow stayed the same.

So - if you ran the outlet of the coolant from the engine into the inlet to your exhaust ex, there would be more heat in your coolant available to be picked up by your domestic ex.
All things being equal - flows - there will be a net benefit to your heat exchange process, I think... if you have the ability to vary the flow on the cool side of your domestic exchanger, you can pick up more heat there, instead of sending hot pre-heated water back to your engine - which may in fact be confusing your thermostat in the configuration you have right now... which might be causing your shutdowns on high temperature...

Here's my reasoning:
Current Setup -

On engine startup, water is slowly warmed in the head, thermostat is closed until the head comes up to temp. There is a small flow through a hole drilled in the thermostat so flow is minimal. Yes?
Head comes up to temp slowly.
At the same time, the exhaust temperature rises very quickly. But, there is very little flow of coolant through the exhaust exchanger, so the water is being heated here to quite a high temp...
When the thermostat opens, flow increases through the cooling system. The first thing the head sees is a slug of high(er) temperature water that has been heated in the exhaust exchanger. This causes the thermostat to behave as if the engine is hotter than it really is... then, as flow increases, and the cooled water that has gone through the domestic exchanger cycles back, the water entering the head is cooler, causing the thermo to close... and the cycle repeats.
I would expect in this system that the cycle repeats in ever-shorter cycles... One of these cycles could introduce water that has been preheated in the exhaust exchanger to the point that when the hot coolant slug hits the high-temp shutdown sensor for the head, it triggers it.

Just guessing here...

With the exhaust exchanger in line right after the head, this situation couldn't happen. You would do most of your heat exchanging where you want it - for domestic uses. You would tailor the fluid flows through the domestic exchanger to control your coolant return temperature. Pull out as much heat as you can here.

In this case, there would be less chance of dropping the outlet temp of your exhaust below 100°C, since the coolant temp would be higher... Of course, this sytem will be limited by the temperature that you can heat your coolant fluid to without boiling it or causing a high-temp failure of something down the flow line.

I could see a lower temperature thermostat being indicated here, say 190°F... keep the engine slightly cooler, pick up more heat in the exhaust ex...

Just some musings from very far away...

[oliver90owner]

my simple comments:

I would go with the coolest water into the exhaust heat recovery as the gas retention time is shortest and the highest delta T is required for the best heat transfer (without condensation problems, of course). 

I would, however, want the engine coolant circulation reduced to a minimum - ie first heat exchanger pretty well adjacent to the engine to reduce the hysteresis in the system, which may well present a problem.

Questions: there is presumably a bypass fitted in this system, so does this circulate through the exhaust heat exchanger and engine?  Is it simply a resticted flow (orifice plate) or what?

A schematic diagram might be useful, but I doubt that would help me to help you unless there are glaring errors!!

Regards, RAB

[veggie]

Jens,

Two things come to mind regarding your pump flow rates....

1] System head:
The taco pump tops out at around 9.8ft. of head, at which point the flow is minimal.
Depending on the size of your coolant lines running the H20 storage tank, I would be surprised if you had less that 8' of head resistance. So the system may be limiting your flow to as low as 1 gpm.

2]Thermostat:
One thing I found with my 6/1 is that the 3 tiny air bleed holes I drilled in the thermostat are enough to cool the engine when I use a pump to force the liquid around. The engine is very slow to warm up (loaded) and the therm doesn't open. I will be replacing the thermostat with a single bleed hole version.
Where am I going with this ?.....
Perhaps your thermostat would not have to open very much to keep the engine cool.
I could be modulating in a partially open position because the incoming coolant temp is low enough that minimal flow is needed. This too would add resistance to you pumps efforts.

Just a few ideas from what little I know of your system.

Veggie

[cognos]

I am beginning to understand.

I now know - there is a variable load on the coolant loop due to demand on the engine, and there is a variable load on the exchange loop, due to the fact that it is possible that heat demand can be met, and the exchanger side heat load can be saturated...
And plenty other variables, like line length, pressure drop, etc...

As far as optimising and controlling this loop, without some pretty extensive control valving and pressure/temperature sensing, I have to change my answer to D: I Don't Know.

This would be easy in my world - but expensive.

[Ronmar]

Jens
   A diagram of your system would be real helpfull... 

As for the plumbing of the exhaust heat exchanger coolant output back into the engine, I would have to say that is probably going to be a problem. If the water entering the engine is already hot, it is going to have a hard time cooling the engine.  I would liken that to plumbing the hot coolant out of one engine into the coolant input of a second engine.  The second engine is going to overheat being fed the heat from the first engine...  If your exhaust heat exchanger is anywhere nearly large enough, it should be putting as much heat into the coolant as the cylinders are(rule of thirds)...  If the engine thermostat is controlling the flow, then when it is closed, it could be allowing the slow flowing coolant to superheat in the exhaust heat exchanger...  What are the temps like at heatex coolant output/engine coolant input?

  To maintain the best deltas, for greatest heat transfer, I would reccomend splitting the flows thru the two heat sources(engine and Exhaust heatex), then re-combining them at the outlets to feed into the home.  A thermostat on the output of both engine and heatex should help balance the flows.  Kind of like this maybe?  the recirculation pipe and valve at the pump will allow you to taylor the pump output to the system.  The open impellar circulation pumps don't have any real issue with being dead headed(when both thermostats are closed), but I never personally like the idea of deadheading a pump...  The recirc valve allows you to control the head pressure that the pump creates, so you don't overpower the system.  basically it is setup when the system is at full load/output, to deliver enough flow to carry away the required ammount of heat.

 

Note the flow thru the heatex.  The coolant flows in the opposite direction as the exhaust gas.  This helps maintain the greatest overall delta between the two fluids and the greatest heat transfer.

As for measuring the flow thru the system, if you know the ammunt of fuel you are burning(energy input), and factor in the rule of thirds, measuring the ammount of temperature rise of coolant passing thru the system should allow you to roughly solve for flow...

[Ronmar]

   A diagram of your system would be real helpfull... 

I probably asked this before but forgot - how do you create these beautiful drawings ?  All I can do is scribble something on paper, take a picture and then post that

Jens

I use the paint program that comes with windows.  They are bitmap images that I save-as a .JPG's for a smaller upload file.  The program is easy to use, but a little slow.  I have quite a library of drawings now, so it is usually pretty easy to cut and paste different bits from other drawings to illustrate a concept.  That last drawing the only thing I drew were the arrows...

[AdeV]

I probably asked this before but forgot - how do you create these beautiful drawings ?  All I can do is scribble something on paper, take a picture and then post that

Over on another board I frequent, that technique is known as Crap-o-Cad, and is widely regarded as the fastest way to outline an idea...

Ron beat me to his reply, so my theory that it was MS Visio (or similar) is blown out of the water. However, that's what I'd use. OpenOffice Draw claims to have the same sort of functions as Viso, with the advantage of being free.

[mobile_bob]

if it were me i would move the heat ex pump to the inlet side, that way there would be no cavitation? maybe?

and the water flow would be more turbulent, which might be of benefit with your exchanger?

just kickin sand with you



btw, have you ever measured the btu recovery rate of your heat exchanger? if so i would be interested in hearing about that.



bob g

[BruceM]

Better check your drawing again, Jens, by my weary eyeballs  your proposed change doesn't do what Ron recommended, and seems to eliminate any input to the cold engine return.   You should be able to reduce this to a single pump, as Ron shows so nicely.

Seems you might also need a cooling loop with radiator, for when your return water temperature gets too high, unless you are satisfied with an auto shut down at that point.  It would be nicer to do it based on return water temp, rather than waiting for the inevitable engine overtemp.

As someone already pointed out,  you have to do a differential pressure measurement at the pump inlet and outlet to be able to then calculate flow rate from the pump specs.  It's the sort of basic info you need to design or tweak something like this.  Veggie may be right- your head may be gone.  Or not- we need more data on pipe length and number of (especially 90 degree elbows and tees) fittings to compute the head loss, even if we have to WAG the heatex head losses.

WAGs (wild ass guess) and TLAR (that looks about right) designs sometimes thump up against reality in an alarming way.