Measuring Air/Fuel Mix In an Oil Burner

In another thread the question of determining optimal air/fuel mixture in an oil burner has been touched upon.

I got to wondering about using an automotive O2 sensor as a means of accurately determining if my fuel /air mix was right as performance of the furnace is highly dependent on this mix. So, I started searching on available automotive O2 sensors following leads where they took me. I got increasingly convinced that a wide-band automotive sensor could be a practical solution. Then I stumbled across this device which is indeed a wide-band automotive O2 sensor coupled with a power supply and Bluetooth connection for a smart phone. It looks like this should work with the only accomadation being a means of cooling a portion of the exhaust gas stream ( could be a simple pipe inserted into the exhaust stream with enough travel to the sensor to ensure cooling of the gases to maybe 400 to 600 degrees F) so as not to burnout the sensor.

https://www.amazon.com/PLX-Devices-...show_all_btm?ie=UTF8&reviewerType=all_reviews

Do folks see factors that might prevent this type of device from being useful? Has anyone any experience with such a device?

160 bucks is a fair bite. But it could reduce guesswork. Looking forward to comments.

 

A search of AA does certainly return a pile of speculative threads and I do not doubt there is useful experience-based information in some thread there, but actually finding it has proven challenging. That may be because threads and their subjects often migrate mid-thread ( I am as certainly guilty of causing this as anyone) making it difficult to actually retrieve useful info. That is one of the reasons I started a new thread with a clear title dedicated to the question in hopes that useful information on the subject would be easily retrievable in the future as a search would surely find it. The other reason was to make sure there weren’t hidden confounding factors I was not considering.

So, if anyone remembers a thread that did not just speculate about the possible use of a wide-band O2 sensor but actually reports on how they worked out in practice, I would really appreciate a link.

If anyone has any experience with actual use, a report pro or con on that experience would be very helpful.

Pointing out theoretical problems which one might encounter would also be helpful.

Denis

 

I think the O2 sensor idea could be a good thing, but I would want to verify it at least once with thermocouples to get a definite relationship established between O2 levels and furnace shell temperature.

 

Thanks, I will look closely at that thread. The problem was not finding a thread it was weeding out the many threads that did not have much useful content.

The modern wideband O2 sensors are generally heated so you don’t have that concern. The one I linked is such a sensor. The attraction for me of the linked sensor is that it is reputedly plug-and-play Bluetooth syncing with your smartphone or pad to provide an information display without having to worry about power supplies, pinouts, and display interfaces. I initially thought I would have to do all that until I found this unit.

 

I do not doubt that there will come a time when I will have become familiar enough with my furnace that I will be able to tell by sight, sound, and smell when it is set at an optimal mix. Until then (and probably even occasionally after that I will want to measure to confirm) I will feel a lot more confident that I am on the right track if I can reliably measure the air/fuel mix. The only down-side I can see to doing so is the expense of the sensor and plugin unit. But unknowingly running the furnace too lean or too rich and at inconsistent settings has big penalties in terms of frustration and time and material wasted.

I agree that the idea of automating control of fuel flow and blower with an electronic control driven by a sensor is a clever thought experiment, but is far too complex and expensive for my tastes and perceived needs. I just want to be able to tune my setup.

Once I have done a bit more reading and listening, I think I will most likely plunk the money for the unit I linked unless there is a better option.

 

when you get to that point your are looking for a 2% O2 reading that should give you. roughly 10% excess air.

Art b

 

Art,

So, the target Lambda is .9 or an air/diesel mix of 13.14 and not 1.0 and 14.6 respectively?


On this page

http://www.plxdevices.com/Wideband-O2-Air-Fuel-Ratio-Sensor-Module-p/897346002726.htm \

For anyone interested there is a decent user manual link which will download the two-page manual.
The link is just under the book symbol. Since the manual is itself a PDF, I cannot give a direct link to it.

 

from my reading i think that you will want a lambda # 1.026 which would be a air fuel ration of 14.8 :1 this is an extrapolation of stoichoimetric ratio of 14.6 : 1 with a 2% oxygen in the flue gas.

these sensors are completely different from what is used in combustion analysis so I'm somewhat guessing. it is possible that my calcs are inverted. if so the lambda # would be .975

Art B

 

Your numbers are in close agreement with those printed in the user manual supplied by the manufacturer of the gauge I am considering. It is helpful to know what number to target for a furnace which is somewhat different from the target of the usual users, hot rod tuners.

 

Couple more melts yesterday using an unmeasured A/F mix but using some of the tips generously offered here. Both a 5 pound and 8 pound melt proceeded nicely with very fluid free-flowing water-like molten iron. With each firing I gain more ability to sense what takes to get high heat. But, though the results have been fairly good, I still don't know if I am operating at optimal efficiency or taking more time and fuel than needed.

I did receive my automotive O2 sensor and Bluetooth interface device late today. I will be anxious to give it a try in the next few days. One challenge I have been mulling over is how to set up a reliable sampling system that will allow me to "sniff" the exhaust but ensure that the exhaust temp is down around 600 C or less at the sensor and that will not melt the intake to the sensor system. Most likely the system will have to be setup to swing into the exhaust and "inhale" for a while and make the reading and then swing back out of the stream. It can make a reading in only a few seconds (or even less), I believe. This system should be relatively blunder-proof so that it does not require babysitting or else there is a risk of becoming distracted and cooking the sensor.

I would welcome any brainstorms you folks might have in regarding design options for the system.

[I accidentally posted this same message initially in another thread on Furnace Break-in. My apologies for the double post]

 

A way to cool the exhaust stream prior to it entering a measuring chamber where the sensor would be located might be to draw the exhaust through a heat sink consisting of a section of pipe filled with steel BB’s. For scale, I am thinking an inch or so diameter and maybe 3 inches or so length. The BB’s would provide high surface area and a fair bit of mass. Such a heat sink would be cheap and easy to fabricate. It would not allow continuous sampling, but that is OK. After a sample is pulled through it, the BB’s could be cooled by pulling environmental air through it. As a means of aspirating exhaust and, subsequently air, I am thinking sucking air using a compressed air stream blown across a relatively small diameter pipe much as an atomizer lifts fluid into the atomizer or use a crude Venturi to do the same. Since the sensor has a very short response time (less than a secondary since my intuition tells me that the heat exchanger would tolerate 10seconds or so aspiration cycles, such a system might work. Other ideas?

Another idea for a heat exchanger could be a plug in a pipe where the plug would be bored with an array of small holes running its length. Just for scale, maybe 1 inch in diameter with twenty 1/16th holes drilled for n inch or so length inmild steel. But, this is more labor intensive to fabricate and is not as easily scaled up as need for adequate cooling.

Lots of idea churning around....

 

W3? built a unit based on perstylatic (sp??) pump with a lambda sensor feedback using a arduino as I recall. It was a very effective furnace and he poured a lot of metal using it for a small commercial venture making CNC kits. Well worth finding the thread if you can.

 

http://www.alloyavenue.com/vb/showt...iphon-burner-setup&highlight=peristaltic+pump is a starting point, W3 started his own site and then gave up casting to start a vape business. I think his own site is down now (unless someone knows better) there was more development there that may be lost now.

 

More dicussion here http://www.alloyavenue.com/vb/showthread.php?890-propane-air-ratio/page2&highlight=lambda+sensor

 

Thank you for the links, Peedee. Sadly, all the images are gone (thank you, Photobucket, for that) so all of the really useful content is also gone.