Exhaust Gas Temperature Testing

[AdeV]

After coming up with a home-made exhaust gas temperature sensor, clearly I needed to test it. The results are interesting, although unfortunately not really scientific enough to warrant being "official" results. The reasons for this will become clear.

Executive Summary (for Policymakers  )

Exhaust gas temperature from a Lister 6/1 with start-o-matic flywheels was measured using a K-type thermocouple, interfaced via a MAX6675 IC to an Arduino microcontroller. The temperature was sampled every second, and fed directly to a display. The data was not archived. The display was monitored while the engine load was varied.

The results were as follows:

- The system quickly equalised at around 120^oC under idle/no load conditions
- Under maximum load, a peak temperature of 310^oC was seen

Introduction

As part of my investigations into a co-gen system, utilising a Lister CS type engine, I was interested in the sorts of temperatures achievable from the exhaust gas. This would also provide a useful training project in interfacing "real world" hardware to an Arduino microcontroller.

For a number of reasons, this test is only a preliminary test - future tests will be much more rigorous - but I present these results now as I find them interesting. Plus, I'd like to query the more knowledgeable here as to whether my results are within expected norms, or whether I've done something wrong.

The problems with this test are as follows:

• The temperature sensor could only easily be mounted around 18" away from the exhaust manifold. For future tests, I will machine a sensor holder which sits right on the manifold
• I cannot consistently load the engine up; there's nothing attached to it. I use a piece of wood to create friction on the flywheel to simulate a load.
• As a result, I cannot produce a load/temperature graph, nor can I guarantee repeatable results.
• As the engine's exhaust terminates indoors, I can only carry out a short run before I must switch off and allow the building to ventilate.
• As a result of that, I cannot correlate exhaust temperature to overall engine temperature.
• Not that I can read the engine (coolant) temperature anyway, having not yet sorted the sensors for these...

So, bearing in mind those limitations, read on...

Equipment

• 1950 Lister 6/1 diesel, with start-o-matic type flywheels
• British "Red" diesel fuel
• A K-type thermocouple, MAX6675 amplifier/A-D converter chip
• "Boarduino" - an Arduino clone board, ideal for breadboard applications.

Method

Firstly, the Arduino circuit was assembled, on some breadboard. The thermocouple was attached and tested for calibration by immersing in boiling water:




A 1.5mm hole was drilled in a section of steel exhaust pipe (NOT in the cast "black pipe" which forms the original Lister equipment - and is, in fact, red). See photo (click the picture for a larger version):



The thermocouple was inserted approximately 3/4" - this depth was chosen as it was the furthest into the hole it would go without serious danger of bendage... The amount of thermocouple in the exhaust stream can be seen from this "after" photo, after the thermocouple was withdrawn:



The thermocouple is located approximately 26" away from the manifold, following the path of the pipe. Obviously, a large quantity of heat was radiated to the atmosphere before the thermocouple, from the un-lagged pipe, and this will serve to deflate all temperatures - by an unknown factor at this time.


Once the equipment was in place and powered up, the cold temperature was noted. The engine was then started, allowed to come up to speed and idle for a few minutes, before a test load was applied. The engine was alternately loaded and unloaded a few times, then shut down.


Results


A start temperature of 16^oC was recorded.

Upon starting, the temperature rapidly rose to around 90^oC, before climbing slowly to around 120^oC as the "black pipe" warmed up. This 120^oC "idle temperature" was noted throughout the remainder of the test.

To simulate a load, a piece of wood was forced against the flywheel to generate friction. Load was applied until the engine was clearly labouring and slowing down, at which time the temperature was taken. This load test was repeated a few times, until the burning wood smoke became a nusciance. The maximum temperature noted was 310^oC.

Between load tests, the exhaust rapidly cooled to 120^oC.

Due to a combination of wood smoke and diesel fumes, the test was then terminated and the engine shut down. Coolant temperature was observed to be "warmish".


Conclusions

• The "on load" temperature was much lower than I was anticipating
• The temperature varies considerably with load; an exhaust heatX would do well to be "tunable", to prevent exhaust gas condensing under light loads

I'm really looking forward to plumbing in a proper external exhaust pipe, adding a controllable load (which doesn't leave carbon all over my flywheels), and making some better temp sensor holders - hopefully, all of this will happen this summer.

Over to you lads - any thoughts on this lot?

[Ronmar]

Jens pretty much hit it, too many unknowns.  There is an old saying in my profession(electronics), "Garbage in = Garbage out".  You should always work from a known to solve an unknown.

Your measured temps are a little below what I see on my 6/1 with 3KW of electric load on the generator.  I stabilize in the mid 600F range, but that is measured with a IR thermometer at the engine exhaust flange.  Why are yours lower?  Again too many variables...  Closer to the head will yield higher temps for the reasons you mentioned.  A higher coolant temp will yield slightly higher exhaust temps.  IF the coolant is only getting "warm" then your engine isn't up to op temp.  A proper hot coolant temp results in a cleaner combustion process and more heat liberated from the fuel to pass out thru coolant and exhaust.  When I did my fuel use measurements, my fuel consumption numbers dropped as engine temp increased. I believe this to be from more complete combustion.

Get the exhaust sorted so you can make sustained runs and get up to a proper operating temp.  Get a controllable load on it, and a way to measure that load.  Get a thermostat on it if you don't already have one, to build and maintain a stable operating temp in the 195F range.  If you are using the same exhaust probe setup, at least wrap the pipe to that point where the thermocouple is sampling.

Good luck.

[AdeV]

Thanks chaps,

I'm aware that the results are - effectively - junk; although I expect the pattern between off-load & on-load will be repeated when I've got the repeatability issues sorted.

Jens - although what you say about the straight exhaust makes sense in "normal" circumstances; I plan on installing a heatex right after the first 90^o bend, it'll run vertically for maybe the same distance as the straight section does now, before angling off to the outside world (if that turns out to be a junk idea - and I would please ask people not to discuss that on this thread, as it's almost irrelevent, except for the point I'm getting around to-  then I can angle it differently); so I will keep a thermocouple in roughly the same place it's at now; as well as have one mounted right in the flange of the exhaust manifold.

I'm more interested in how much heat I am likely to be able to harvest - and how much control needs to go into the HeatX coolant flow - depending on load.

[mobile_bob]

Ade:

this will get you close

for instance if your engine consumed 0.6 gallon per hour at full rated load, you can assume that 1/3 of that will go
out the exhaust or about .2 gallon.  figureing that a gallon of pump diesel has about 138kbtu's, then

135k x .2 gallon = ~27kbtu's, but your exchanger will likely be about 75% efficient so, you get

27kbtu x .75 = ~20kbtu

so that works out to about 20kbtu at full load with an exchanger that can harvest without degradation due to carbon
buildup over time, so maybe you can figure on a bit less?

the problem becomes how much can you harvest at 3/4 load, 1/2 load, and 1/4 load?

heat exchangers do pretty well in a narrow range of operation and drop off sometimes dramatically either side of optimum
so you size according to your average load and give up a bit on either side of that load level.

ymmv, but i think you get the idea

otherwise i think you are on the right track, and you will likely find that the exhaust gas is around 700 degrees F at full rated load
that would be my guess with a prob right in the gas flow, and i would not expect a dramatic difference  until you move out a bit further.

bob g

btw, i like your presentation style!

[Ronmar]

I'm more interested in how much heat I am likely to be able to harvest - and how much control needs to go into the HeatX coolant flow - depending on load.

Well those two questions seem pretty simple to me:). How much heat are you going to be able to harvest from the exhaust?  Probably pretty close to 1/3 the BTU content of the fuel being consumed.  What sort of coolant flow control?  Why not control the flow the same way they do in an engine?  Put an automotive style thermostat on the outlet of the ex heatex.  Let it modulate the flow to maintain a desired heatex outlet temp.   

[AdeV]

Bob - I was trying to remember back into the mists of time, the sort of layout we used in school for physics & chemistry experiments... not sure I got it totally accurate, but it's near enough I think. OK, I admit that I applied scientific headings to a totally unscientific test.... but what the heck, it kept me off the streets for a couple of hours.

Ron; That sounds like a sensible "low-tech" way of going about it... I'd thought about using the Arduino to run variable-speed pumps & open/close valves... much more "Heath Robinson" (I believe the US equivalent is Rube Goldberg) and ultimately more rewarding, but probably less effective and more failure prone....

[Ronmar]

"low tech"  That is my middle name:)

[DRDEATH]

I thought one of the rules on this forum was not to ATTACK each other.  AdeV stated his experiment was not completly scientific. I believe Jens post could have been a little bit more diplomatic. Just my thoughts. When I have an idea or I feel that I have something to share I will keep it to myself. DD

[vdubnut62]

Ease up DD! They know each other well enough to make that exchange, and as you will notice no offense meant or taken. We are not a thin-skinned bunch here,
nor are we Politically Correct. (usually!) What you mistake for an attack is a simple exchange among friends.
Ron

[LowGear]

Aloha AdeV,

Nice to know information.  I appreciate your efforts and applaud your transparent research model.  As a non member of the Rocket Surgery and Brain Science foundation I can say you have far exceeded my efforts and your data will be uploaded to my interdynamicgalactic navigation system ASAP.  I hope to return by Earth's 30th century.  Just me and the bugs. 

In the mean time I hope to build a bean and coffee cooker on the exhaust of my Witte.  Your data support my visualizations completely.  Much Mahalo.

Cheers,

Casey

[AdeV]

Aloha AdeV,

Nice to know information.  I appreciate your efforts and applaud your transparent research model.  As a non member of the Rocket Surgery and Brain Science foundation I can say you have far exceeded my efforts and your data will be uploaded to my interdynamicgalactic navigation system ASAP.  I hope to return by Earth's 30th century.  Just me and the bugs.  

In the mean time I hope to build a bean and coffee cooker on the exhaust of my Witte.  Your data support my visualizations completely.  Much Mahalo.

Aloha to you too Casey... I have almost exactly no idea what you said above, but if in some small way I managed to help, then I have truly exceeded my potential. Well, probably anyway. And while we're at it, could you do me a favour & leave the bugs wherever it is that they are just now? It's bad enough knowing that one day we're going to have to kowtow to the cockroaches, without yet another set of bug overlords to get in the way...

Meanwhile.... there has been no further progress on this project... various other projects & factors have conspired to keep the Lister silent so far. Then again, the summer is not yet done (Hah! Tell that to the weather - it's been positively autumnal for the last 2 weeks), and a £9000 electricity bill means the Lister may actually have to pull its own weight soon...


I don't know what my average and peak office electricity consumption is (other than I do know that when the air con kicks in, the UPS twitches); does anyone know if there's a UK equivalent of the Kill-a-Watt meter, that doesn't cost stupid money?

[AdeV]

£9k does cover quite a few years when they've either forgotten to bill us, or billed the wrong people... it's all a monumental clusterf**k, and the easiest way to sort it might be to tell them to go to hell & I'll generate my own electrickery...

[DRDEATH]

Well I guess the mud is in my face. In another forum there was to much of this and thats why I came over. I stsnd corrected. DD

[AdeV]

Well I guess the mud is in my face. In another forum there was to much of this and thats why I came over. I stsnd corrected. DD

I don't see any mud, on your face or anywhere for that matter... I knew up front that my experiment wasn't exactly scientific; and yes, it was still painful for Jens to point that out, but not fatally so...

I speak only for myself, but I'd much rather you hung around here than not: So long as you keep away from the politics, this has proven to be a genuinely friendly forum; I can only think of one other (which is nothing to do with engines) which matches this place for lack of animosity.

[LowGear]

Hi AdeV,

Aloha to you too Casey... I have almost exactly no idea what you said above

I often feel this is when people understand me the best.  Yes, there was a bit of sarcasm in my message but firing up the Witte and enjoying hot coffee and beans for breakfast is one of my dreams.  Hopefully before I turn 70.  I hear folks over seventy can have gas.  Once I turn seventy then I can redirect my gas back into the Witte.  This Co-Gen stuff has really sparked my imagination (Oops, did I write spark and personal gas in the same paragraph?)

Nice contraption you rigged up to do the data gathering as well.  My electronics hasn't moved much past the superhyrodon (it has been 45 years now hasn't it) thingie we built in the Army training program.

Casey