Coolant loop flow measuring

[Ronmar]

I don't quite understand your change either Jens.  Are you adding a second pump?  If so, the basic issue I got from your initial description remains.  The coolant supply to the engine is still from a heat source(ex heatex).  If this is a pump plumbing change, and not an addition, then as Bruce noted, where is the engine getting it's coolant supply from?

[billswan]

Jens

You probably should install a gauge on each side of the pump you are now using to give the flow determination a shot. That should not cost to much.

My vote is to install some sort of bypass so the coolant is moving all the time and then when the thermostat opens it dumps heat to the house until the house can take no more then the over temp murphy shuts down the listeroid. But the pump keeps moving heat at least until thermo closes.

Billswan

[cognos]

Here's how I think I would do it.

I would have the exhaust heat exchanger be a fixed flow loop, always full flow. It could be adjusted with a bit of back pressure with a ball valve on the coolant inlet, to provide a minimum coolant flow through the engine, through the hole drilled in the thermostat.

The engine coolant loop would be on thermostat; it would open/close on engine cooling demand. The flow through this loop could be fine-tuned so that with the thermostat fully opened - maximum cooling demand from the engine - there would be adequate flow 'robbed' from the constant flow loop[ to satisfy demand.
This loop would join the constant flow loop just before the domestic water exchanger(s).

Since both loops dump their heat in the domestic heat exchanger, then return back to the coolant pump, there will always be cooled coolant available to the engine - and maximum heat would always be picked up in the constant-flow exhaust exchanger loop.

Just my take on it, I think others are recommending something similar that would work, too.

[BruceM]

Cognos's proposed configuration is the same as Ron's proposed fix, both returning the coolest water (storage heatex)  to the engine.  It's the only configuration that makes any sense to me.

Jen's existing setup, which returns exhaust heatex hot water to the cold return to engine is...not going to provide the best heat transfer efficiency, and will have serious problems as storage temps rise, as the engine sees "cold" return water temps of storage temp PLUS the rise in temperature from the exhaust heatex . 

Some flow rate data and/or data on pipe runs and fittings would show whether adequate flow is being provided. 

There is no way this system is performing at anywhere near optimum, as is. 

[Ronmar]

Well since the engine was providing 220F or less, and the exhaust is providing 400F or more under load, I would hazard a guess that you are making steam in the exhaust heat exchanger...  Your temp measurements between exhaust heatex and engine should show this.

Splitting the loops and putting a 90C thermostat on the output of the ex heatex, would allow the ex heatex to regulate it's own temp and stay below boiling.  That of course is dependent on the pump being able to provide enough flow to transfer enough BTU away from the heatex...

[Ronmar]

Oops, missed the secondary loop part.  How is that loop circulated?  Got a drawing of that side?  If it was circulating, and the shutdown limit out of the cogen is 220F, the secondary shouldn't have boiled unless it was completely heat saturated...  I think you mentioned that the secondary loop provides ALL your engine cooling?  If so, I would think you would have hit engine shutdown from lack of cooling long before then..

A little dab of heat sync compound where your sensor contacts the pex might help with your temp measurements...

[quinnf]

It might be worthwhile to immerse the thermocouple in boiling water and an ice bath to verify their accuracy.

Quinn

[dubbleUJay]

Maybe I am seeing thermal lag and that is all (the thermocouple has a very tiny thermal mass compared to the DS18S20).
Jens

Jens, I think that's exactly what's happening! I'm using LM35 for sensors rather than the 1-wires, but the packaging is essentially the same. I'm also seeing a lag in temperature readings.

[mobile_bob]

could you put the sensor into a piece of copper tubing?

crimp one end shut, solder it off to seal it
drop the sensor down in and epoxy fill it, or pot it with whatever you like

then fit the copper tube into a ferrule T fitting?

that would get the prob down into the flow, and the ferrule would give you positive sealing

what i am thinking of using anyway.

bob g

[dubbleUJay]

Jens, maybe youve seen this:
http://www.instructables.com/id/Waterproof-a-LM35-Temperature-Sensor/

I didn't do it exactly like that, but it gave me the general idea.
I also used a RG59 crimping tool to seal the wire end off, filled it with heat compound and epoxy the open end closed.
Two of them that will screw in, I just soldered the copper tubing into NPT threaded fittings before fitting the sensor and wire.
Mine now looks like the automotive ones more or less.

[BruceM]

I like your method, WJ,  more than the link's method of using PVC and copper, and silicone.

The "instructables" article has a guy crimping the end of the tube in a vice- with the LM35 in place???!!!
Wow.  And  silicone that can never cure in a tube with wires.  Hmm. 

Steel filled epoxy is decent for thermal conduction, and a decent permanent plug in many fluids.

[dubbleUJay]

Once I land back on my feet after all the stuff I'm busy with at the moment, I'll write-up a small paper with pictures of how I did it, it's quite simple though, the crimping tool does a great job.
Afterward I used a short piece of heat-shrink tubing over the wire-end, just to neaten it up.

[billswan]

Jens

A comment and then a question.

In your quote below you say "the exhaust heat exchanger seems to be recovering significantly more BTU's than the engine coolant" gee that seems to fly in the face of the old rule of thirds, 1 third power 1 third heat into radiator and last 1 third  out the exhaust as lost heat. Not arguing with you just wondering out loud? Wounder if any smart guys here would comment on that? Oh and please excuse me if I am not following what you wrote.

You also say "This unit is outside the engine enclosure and not yet insulated" I thought it was inside the enclosure? We are talking about the exhaust heat exchanger aren't we or am I missing something ?

Billswan

The one thing that has me puzzled is the fact that given the same coolant flow rate, the exhaust heat exchanger seems to be recovering significantly more BTU's than the engine coolant. This unit is outside the engine enclosure and not yet insulated and I expected considerably less output from it. Some plumbing changes in the coolant system will be in order.

Jens

[quinnf]

Re: the unexpectedly low heat value of the cooling water, maybe thermal resistance is the culprit.  Heat transmission through relatively thick cast iron (cylinder head) and steel (cylinder liner) is going to be slower than through the thinner sections in the exhaust gas heat exchanger.  It might be that the heat is being scavenged away by the coolant faster than it can pass through the metal in the cylinder head and cylinder liner, whereas in an exhaust gas heat exchanger you have thinner metal sections for the heat to pass through.

If the t-stat is cycling, it sounds like maybe you need some flow restriction.  Consider closing a valve just enough to keep the thermostat fully open at max load.

Quinn

[mobile_bob]

my .02 and its worth about that,

i would  not want the Tstat cycling, i would want the engine up to temp and have the tstat running open
thermal cycling in the engine rarely is a good thing.

heavy trucks run with 195 tstats, and the engine runs over 195 to about 205, with the fan used to moderate
the temps and keep things running at a fairly stable temp under load.

now i suppose one could moderate the temp in much the same manner by regulating coolant flow, so that the tstat
could find its equalibrium point and run at some opening and stay pretty much there at that load.

i will defer to fellars on the virtues of coolant over exhaust heat flows, but it has been my observation that exhaust heat
being of higher quality seems more efficient at transfer to the coolant medium, if it is because of thinner wall sections or higher
heat differential or both, i don't know for sure,
but i suspect a bit of both myself

bob g