CR4® - The Engineer's Place for News and Discussion®

 Previous in Forum: Elevator Design Next in Forum: How to Calculate Water Velocity
Member

Join Date: May 2012
Posts: 9

# Boiler Flue Gas Temperature Problem

05/06/2012 1:55 PM

We have HFO (Oil fired boiler) that generate steam for 300MW Steam turbine. Boiler Steam generation capacity is 1000 tons/hr.

At Gas Air Heater (GAH) outlet flue gas path, there are thermocouples which read flue gas temperature called 'Back End Temperature'. At this point, 150C temperature is observed usually.

Then this flue gas is passed through the duct (30 meter length) to stack. Near stack, thermocouples also mounted there which read flue gas temperature. At this point, 160C temperature is observed.

Now our concern is WHY THIS 10C TEMPERATURE IS INCREASING BETWEEN 'BACK END TEMPERATURE' & 'STACK INLET FLUE GAS TEMPERATURE' - KEEPING IN MIND THERE IS NO INGRESS OF ENERGY BETWEEN THESE POINTS? According to the system design, temperature should be dropped here rather than increasing!

We have 4 boilers & same behaviour is observed on all 4 boilers. We have also ensured the integrity of our temperature sensors & they all are OK. Thermocouples type & mounting position is same at both sensing points.

Thanks

Interested in this topic? By joining CR4 you can "subscribe" to

Check out these comments that don't yet have enough votes to be "official" good answers and, if you agree with them, rate them!
Guru

Join Date: Aug 2008
Location: Florida & Ireland
Posts: 1024
#1

### Re: Boiler Flue Gas temperature Problem

05/06/2012 4:21 PM

As you state "there is no ingress of energy between these points" - therefore it must come from within. The only explanation that I can think of, is that there is an exothermic chemical reaction taking place in the flue gases that is favored by the lower temperature and, perhaps, the lower pressure near the stack. Some exothermic chemical reaction which results in more gas molecules in the products than in the reactants, would illustrate Le Chatelier's Principle in action i.e. the system tries to oppose the constraint (drop in pressure) by producing more molecules.

Since your gaseous exhaust is going to consist mainly of CO2, H2O, N2 and some O2, with traces of NOx, CO, Ar etc., it's hard to think what might be reacting.

One further thought: Could there be metallic or oxide deposits in the flue that might catalyse some gaseous reaction?

Member

Join Date: May 2012
Posts: 9
#4

### Re: Boiler Flue Gas temperature Problem

05/07/2012 1:46 PM

I just want to inquire further, in flue gases we have major contribution of CO2, SOx & NOx gases as we are burning Heavy Furnace Oil in the Boiler.
So do you think such SOx & NOx gases can contribute in exothermic reaction ?

Also, at point where 150C temperature is observed (say Point"A") it is close to furnace and at that point where temperature get 160C it is far away from furnace(say point "B"). So is there any provision that flue gases at Point "A" may not get properly mix b/c its close to furnace & when flue gases reach at Point "B" its temperature increases by 10C after mixing?

Guru

Join Date: Aug 2008
Location: Florida & Ireland
Posts: 1024
#6

### Re: Boiler Flue Gas temperature Problem

05/07/2012 2:21 PM

Your comment on mixing sounds very feasible. I wouldn't have thought that there was enough SOx or NOx in the flue gases to support any chemical reaction that would raise the entire volume of gas through 10 deg C.

Commentator

Join Date: Aug 2010
Location: VA
Posts: 78
#2

### Re: Boiler Flue Gas temperature Problem

05/06/2012 6:40 PM

Maybe your probes are reversed, and its a 10 degree drop? Woody

__________________
Life is short, eat your dessert first .......In theory there is no difference between theory and practice. In practice there is. Yogi Berra
Guru

Join Date: Aug 2008
Location: Florida & Ireland
Posts: 1024
#3

### Re: Boiler Flue Gas temperature Problem

05/06/2012 9:41 PM

I hope you are right. I don't care for my explanation - clutching at straws methinks.

Member

Join Date: May 2012
Posts: 9
#5

### Re: Boiler Flue Gas temperature Problem

05/07/2012 1:48 PM

I have already checked this option. Probes are not reversed.

Thanks

Commentator

Join Date: Aug 2010
Location: VA
Posts: 78
#7

### Re: Boiler Flue Gas Temperature Problem

05/07/2012 3:10 PM

When in doubt, check. I would consider swapping the probe positions to insure that they are not skewed. Since they are the same the test is a simple one and the results would give a solid result one way or the other.

I believe that the temperatures would drop due to losses and that the gain is incorrect for one reason or another. Multipath grounding or ground loop issues could also throw off the readings.

To beat this to death, you could move the probes together and see if they agree, or else publish the results as a new source of heat energy. Woody There is a skunk in here somewhere.

__________________
Life is short, eat your dessert first .......In theory there is no difference between theory and practice. In practice there is. Yogi Berra
2
Associate

Join Date: Apr 2012
Location: Lollipop Land
Posts: 34
#8

### Re: Boiler Flue Gas Temperature Problem

05/07/2012 4:15 PM

There may be some physics or chemistry explanation, but given the number of temperature measurment installations I've seen that have problems, I tend to lean towards measurement error.

1-3°C errors are not all uncommon due to ignoring CJ error, stated temp coefficients of the instruments and/or sensor, expected error within the published limit-of-error sensor specification (which is much, much broader than most people realize.)

For example, Type K standard limit-of-error is ±2.2C. So, an out-of-the box, brand new, Type K thermocouple at the extreme of the limit, +2.2°C, combined with a +1.0°C CJ error in the same direction has a certainty of a 3.2°C 'uncertainty'.

It could be that the downstream temperature appears hotter because the sensor location 'sees' a lower temperature. The upstream point might have some installation factor such that the thermocouple doesn't see the higher temperature, like when the element is in a thermowell. Inserting a sheathed sensor into thermowell where the sheath does not bottom out can causes error due to both the air gap and a location not reflected by the length of the well.

I've witnessed 4" long sheaths inserted into 12" thermowell bores "because they fit and that's what we had available". The measurment point was right at the wall with NO insertion into the process stream. No wonder it was not reflecting the temperature of sensor inserted into the flow stream.

Other disasters I've seen that create an error in thermocouple temperature readings:

- Configuration error in the transmitter/PLC/DCS/controller/recorder/indicator. Primarily mismatch between the transmitter's analog output scale and the scale assigned to the 4-20mA in the receiver. transmitter is scaled 0-200°; input is scaled 0-250°C. My bet's on this one. VERY common.

- multiple menu locations for offsets or slope adjustments on thermocouple inputs. I know of one transmitter/PLC/DCS/controller/recorder/indicator device that has four, count 'em, four different configuration menu locations where an offset to a thermocouple reading can be applied. Do NOT underestimate the likelihood that one tech adds 0.5° in one setting, another adds 0.6° in another, and 3rd got it wrong and did a slope correction of 1.1, rather than a 1.1°offset. It ain't hard to get 10°C error with a 10% slope correction.

- measuring the wrong points. Well, the points are temperatures, but they're not the temperature the labels say they are.

- Arbitrary assignment of a (external) cold junction compensation value by the receiver (AI card in transmitter/PLC/DCS/controller/recorder/indicator) rather than actual CJ measurement and correction factor. (we didn't set it up that way. The mid shift guy must have done that)

- Thermocouple CJ compensation not used (look up table used on millivolt reading, no CJ comp); measurement of 'hot' points read lower in temperature than expected (but typically more than 10°C error).

- Wrong type extension wire

- Reversed polarity connection of extension wire at junction boxes or at the thermocouple head.

- Substitution of copper wire rather than extension wire creating false junctions and a 'remote' cold junction location.

- Use of standard terminal blocks in intermediate junction boxes rather than isothermal blocks

- Chimney effect of vertically oriented terminal blocks creating gradients across (+) and (-) terminals.

- Wrong type thermocouple plug and jack for the type wire/thermocouple in use.

- Heat sink on one T/C terminal (CJ sensor on that terminal introduced significant error due to panel fan blowing on the terminal strip)

- the gamma radiation effect. But in a HFO burner ? ? ? Is this site have direct exposure or is it in the fall-out path of Fukushima, by any chance?

- the Virtual Junction Error when the temperature along length of the T/C exceeds that of the hot tip. A secondary or virtual junction is created with a resulting error.

__________________
Toto, I've a feeling we're not in Kansas anymore
Member

Join Date: May 2012
Posts: 9
#10

### Re: Boiler Flue Gas Temperature Problem

05/13/2012 1:39 PM

We have 2 thermocouples installed (vertically mounted) in the center of flue gas duct in parallel. Its of T-type. There is one junction box in the field in which these two thermocouple gets in parallel to have average output value. Then this thermocouple wire terminates at thermocouple input card of DCS (located 100m away). CJ compensation is at the input card of DCS. There is no Copper wire in between. Whole wire comprises of extension wire.

I have conducted few experiments whose observations are as under:

1) I have measured temperature at the head of both thermocouple. I found 160 C at both thermocouple head. Then I measured temperature at input & output of averaging circuit . At both points, I also found reading approximately 160 C.
This (160 C) reading is also appearing at the DCS.
Hence it means there is no issue in the filed wiring / Avg circuit / input card / CJ compensation etc.

2)About 50 meters UP STREAM of the aforementioned pair of thermocouple, there are 3 thermocouples installed in the same topology. Same experiments have also been conducted at this point to ensure integrity of the TC & field wiring. This TC reads about 150 C.

Now the question is, why the temperature is raising 10 C at the down stream of the point where temperature is 150C. Logically, the temperature should be dropped rather then increasing ??

Looking forward for your expert opinion.

Guru

Join Date: Jul 2010
Posts: 651
#9

### Re: Boiler Flue Gas Temperature Problem

05/09/2012 10:50 AM

I, too, vote for measurement error.

I laughed so hard I spit my coffee out when reading MagicWand's list of measurement error possibilities. How true. I see the same thing over and over.

Anyhow, a case-in-point showed up on control.com a day or two ago.

Different T/C's than the receiver was configured for. Believe me, it happens.

Member

Join Date: May 2012
Posts: 9
#11

### Re: Boiler Flue Gas Temperature Problem

05/13/2012 1:40 PM

We have 2 thermocouples installed (vertically mounted) in the center of flue gas duct in parallel. Its of T-type. There is one junction box in the field in which these two thermocouple gets in parallel to have average output value. Then this thermocouple wire terminates at thermocouple input card of DCS (located 100m away). CJ compensation is at the input card of DCS. There is no Copper wire in between. Whole wire comprises of extension wire.

I have conducted few experiments whose observations are as under:

1) I have measured temperature at the head of both thermocouple. I found 160 C at both thermocouple head. Then I measured temperature at input & output of averaging circuit . At both points, I also found reading approximately 160 C.
This (160 C) reading is also appearing at the DCS.
Hence it means there is no issue in the filed wiring / Avg circuit / input card / CJ compensation etc.

2)About 50 meters UP STREAM of the aforementioned pair of thermocouple, there are 3 thermocouples installed in the same topology. Same experiments have also been conducted at this point to ensure integrity of the TC & field wiring. This TC reads about 150 C.

Now the question is, why the temperature is raising 10 C at the down stream of the point where temperature is 150C. Logically, the temperature should be dropped rather then increasing ??

Looking forward for your expert opinion.

Guru

Join Date: Aug 2008
Location: Florida & Ireland
Posts: 1024
#12

### Re: Boiler Flue Gas Temperature Problem

05/13/2012 3:50 PM

Harking back to my earlier comment, could it be that the oxidation of your hydrocarbons is not quite complete at the 150 C measurement point, perhaps due to inadequate mixing, or some heavy non-volatile component in your fuel is still undergoing oxidation, resulting in a 10 C rise?

Associate

Join Date: Apr 2012
Location: Lollipop Land
Posts: 34
#13

### Re: Boiler Flue Gas Temperature Problem

05/17/2012 7:48 PM

The checks you've done at the sensors themselves are very valid checks, ruling out most of the wiring, receiver scaling and cold junction error issues.

There are two things that I can think of that might contribute to a measurement error
- length, if thermcouple sheath is inserted into a thermowell

Measurements where thermocouples do not bottom out are problematic (lag and value offsets).
If the thermocouples are inserted into a thermowell, I'd suggest removing a thermocouple and inserting a stiff wire to see where the bottom of the thermowell is and comparing the length needed to bottom out with an actual thermocouple sheath length.

- length and position in the duct
Apparently the thermocouples are top mounted (you mentioned vertical orientation). And multiple thermocouples are paralleled for averaging. That suggests an effort to get an average temperature across the duct.

The one thing that might be worth checking is to see if the upstream thermocouple length and subsequent position in the duct is what is expected, as compared to what is shown on the drawing. If the length is too long, the sensing point might be showing a cooler level in a thermal gradient because the temperature is lower "close to the floor".

If those installation points can be ruled out as a source of error, then you likely do have a 10° differential.

Combustion is not my field, but I strongly suspect that, if, as suggested, residual fuel combustion continues in the exhaust flow stream, releasing more heat, then an upstream/downstream test with a combustion gas analyzer would reveal lower oxygen concentration and higher CO2 concentration downstream as a result of combustion between those points. Maybe a combustion guy can chime in and offer a more educated opinion.

__________________
Toto, I've a feeling we're not in Kansas anymore
Associate

Join Date: Apr 2012
Location: Lollipop Land
Posts: 34
#14

### Re: Boiler Flue Gas Temperature Problem

11/16/2012 8:32 PM

Was there ever any resolution to the temperature measurement issues?

__________________
Toto, I've a feeling we're not in Kansas anymore
Guru

Join Date: May 2008
Posts: 1798
#15

### Re: Boiler Flue Gas Temperature Problem

01/25/2013 4:16 AM

Dear Mr. evil.genius

My doubt is - some amount of combustible matter escapes and continues to burn and releases heat energy - resulting in Rise in Temp. of the flue gas.

Pl. check for this, by analysing the gas before and after the duct referred by you.

Thanks,

DHAYANANDHAN.S

Interested in this topic? By joining CR4 you can "subscribe" to