INTELLIGENT WORK FORUMS FOR ENGINEERING PROFESSIONALS
Come Join Us!
Are you an Engineering professional? Join EngTips now!
 Talk With Other Members
 Be Notified Of Responses
To Your Posts
 Keyword Search
 OneClick Access To Your
Favorite Forums
 Automated Signatures
On Your Posts
 Best Of All, It's Free!
*EngTips's functionality depends on members receiving email. By joining you are opting in to receive email.
Posting Guidelines
Promoting, selling, recruiting, coursework and thesis posting is forbidden.

Boiler efficiency calculations(4)

Is there a source that provides detailed boiler efficiency calculations that shows how a boiler efficiency value is derived from the standpoint of the combustion side data. This gets most confusing when the values are intermingled to suit whoever the seller is! e.g. a firetube boiler fired on natural gas has CO2 of 10.5% with stack temp. of 220 degrees celcius at full load and ambient of 30C. The fuel has a higher calorific value of 40MJ/m3 and a lower calorific value of 34 MJ/kg at stp. (The stoichiometric CO2 is 11.9% and the stoichiometric air fuel ratio is 9.26) We have a formula such as 100(K * stack temp(C)  air temp(C))/CO2(%) where K = 0.54 so the above equation looks like this =100(22030)*0.54/10.5 = 90.23% boiler efficiency. It is apparently based on the 'Hassanstad' formula(?) So, 1) where does the K factor originate? and 2) where does the fuel gross or net calorific value finds its way in? and 3) what does the efficiency mean in practice? e.g. if we burned 100kg of the above fuel of 40MJ/kg = 4GJ heat in the furnace  how much of that find its way in to the water / steam? (90.23% ?  with the balance being e.g. stack losses, radiation, latent heat of water vapour?) 

TBP (Mechanical) 
7 Jan 05 9:52 
If the only information you have is fireside data, then you can calculate combustion eff. And while that's important, it's not boiler eff. To calculate boiler eff, you need to determine the net heat out, and divide it by the heat input for the same time period. The calculation itself is dead simple. Acquiring accurate data is another matter altogether. Most boilers have, at best, seriously inaccurate steam meters. Many have none at all. Many plants have one fuel meter for the whole facility. Others have one meter for the entire boiler room. Few have accurate meters for the fuel input to each individual boiler.
I'm really happy to see someone (else) is looking at this with a critical eye. Even among socalled "heating professionals", the confusion regarding "boiler eff", "combustion eff", "AFUE eff" is something approaching total. Manufacturer's very much tend to pick the one that suits their purpose. Everyone should scrutinize things like efficiency numbers that are presented, and ask just how this value was calculated. Figures don't lie, but liars can figure. 

25362 (Chemical) 
7 Jan 05 11:01 

Thanks TBP! I have figured the combustion efficiency angle but do not see the realworld relevance of it when one knows what stoichiometric is and corresponding oxygen and CO2! I got suckered nicely when that was put forward as the 'efficency' but without the rider of 'combustion'.
We have measured the water flow at inlet and the temp. corresponding to each value and have the gas supplier actual billing data of GJ and m3. By passing the data through steam tables (error in taking averages for feedwater temp.  blowdown is heated to steam temp. so that counts for my calculation) we are able to do an excel goalseek to get the efficiency that meets the steam heat load from the steam tables  knowing the gas energy on the other side of the equation. We get say 85%  86%. (Stack oxygen levels 3.3% at 20% load and under 2% oxygen at full load  tough to get better combustion there! 3 pass boiler  220C boiler stack exit temp.)
That's nice, but it is history  it has already happened and you have been suckered in to buying the boiler that now does not perform to the 'efficiency' quoted!
Is a 90% gas boiler efficiency realistic? (No economizer or air preheat)
Does anyone have a gas boiler (of reasonable size) that operates up there?
Oh well... 

quark (Mechanical) 
7 Jan 05 12:09 
This Thread404100205 may be of some help to you. As already stated by TBP, these calculations give you a ballpark figure but the best way is to check it by the ratio of heat output to heat input. We have a 3ton/hr, 10kg/cm ^{2} boiler with 220 ^{0}C stack temperature without air preheater and economizer. It is an oil fired boiler. The best efficiency I got was 85%, the day after the fire tubes and shell side were cleaned thoroughly. 

TBP (Mechanical) 
7 Jan 05 12:54 
With a properly sized, installed and calibrated steam meter, equiped with pressure compensation, and an accurate fuel flow meter, the best eff I've ever seen on industrial scale boilers in the field was 82%. That's a proper boiler eff based on the BTU of the steam out, minus the heat of the feedwater (which in most industrial plants should be pretty constant), divided by the BTu in the fuel over the same time period. This is the "big picture". It captures all of the losses, but doesn't identify where they are. Other calculations, like combustion eff, are important to determine where those losses are occuring. Remember that equipment like air preheaters and economizers, while the offer valuable heat recovery and help with overall plant eff, do NOT have any effect on boiler eff. Any heat that provides the boiler with a "running start" must be discounted before the boiler eff calc is made.
This same mass flow  BTU calc can be done on anything that's instrumented properly  gas fired unit heaters, hot water boilers, forced air furnaces, etc. But they're not. The eff numbers tossed around by most manufacturers of this type of equipment are the ones that get them closest to 100%, regardless of the calculation method used. Steam equipment typically suffers badly by comparison for playing by the rules of engineering, as opposed to the rules of marketing. 

25362 (Chemical) 
8 Jan 05 1:49 
The % design load (turn down), and weather conditions that may influence losses, should be considered because they are supposed to affect the efficiency and the overall economics of the boiler system. 

You could try one of the acceptance codes if you want a detailed procedure. There is ASME PTC4, BS2885, DIN1942
athomas236 

TBP (Mechanical) 
8 Jan 05 10:41 
25362  We found, as did a sister plant, that boiler eff would run in the high 70's to low 80's percent range  maybe a 5% spread  spanning any load high enough to keep the boiler on automatic firing. Something that kind of surprised us, was that at loads over about 75 or 80% of boiler capacity, the eff actually started to drop back a little bit maybe from 82% to 80 or 81%. Note that we used the LHV of the fuel (natural gas), as this was the corporate standard. 

TBP  1. How would the boiler efficiency compare if the higher heating value of the gas was used in the calculation? 2. Also, you mention the eff. dropped off a little as boiler load moved above 75 or 80%. That is interesting  do you have data for a spread of loads  e.g. oxygen %, stack temp, ambient temp., feedwater temp? (The oxygen gives me an idea of combustion efficiency). 3. what was the boiler maker`s efficiency rating for the boiler? (and based on what parameters?)
My data shows the combustion efficiency to drop off as load increases  from 85% comb. eff. at 20% load (& 3.3% oxygen) to 82.5% at full load with 1.8% oxygen)  this seems odd. quark  1. was the 85% efficiency based on higher heating value or lower heating value of the fuel? The formulae in the thread were interesting  just need to understand where the empirical constants are derived from. 2. what was the boiler maker`s efficiency rating for the boiler? (and based on what parameters?)
25362  you are correct in the turn down and ambient conditions influencing the boiler efficiency. Turndown is managed in my mind by the oxygen level in the flue gas. If that can be held as close to 2% as possible throughout the range its influence is managed. 2% being for the 10% excess air to ensure complete combustion  but if CO is zero at the furnace exit, there is no chance of CO ignition in the reverse chamber and the oxygen level can be pulled back a fraction.(?)


quark in the thread above you posted an equation for HFO:
%Heat lost in moisture in flue gas = [[212T1+970+0.5(T2212)](9H+M)]/GCV
M is % moisture in fuel H is % hydrogen in fuel T2 = Flue gas temperature in deg.F T1 = Air inlet temperature in deg.F
1. What units is the GCV expressed in? 2. Is this valid for any fuel?


TBP (Mechanical) 
9 Jan 05 11:51 
toohotforme  I can't remember all of the data, and it's been a few years since I worked in that plant. Feedwater temp was 227*F. There may have been some minor excursions, but they would have been so short lived that they would have been insignificant. HHV of fuel will give you a lower boiler eff value, because you're dividing the neat heat output from the boiler by a larger number.
Flue gas temps increase with boiler load. We normally fired gas, but were on interuptible service, so the highest loads we saw were when we were on #2 oil. The fuel gas temps at near full load on 40,0000 and 60,000 #/hr boilers firing to maintain a header pressure of 125 PSIG) would be in the 590*F range, and pushing 600*F. Note that since we were firing to maintain header and not boiler pressure, the drum pressures would be 140  145 PSIG at full load. (This is why you need pressure compensated steam meters.)
The boilers I'm referring were 20 years old when I got to them, so I never saw the actual manufacturers performance sheet. Manufacturer's of boilers at least used to publish performance data, and outline specific conditions, external to the boiler, required to hit this mark. These included minimum feedwater temp, combustion intake air temp, a maximum continuous blow down rate value, a specified BTU value for whatever the fuel was, etc. These would be something like 220*F feedwater temp, 60*F intake air temp, and a max CBD rate of (maybe) 5%.
Overall, if you've got an industrial steam boiler that's hitting 80% BOILER eff in actual field service, that's about as good as it gets. 

25362 (Chemical) 
9 Jan 05 12:55 

msonnet (Electrical) 
10 Jan 05 7:47 
I did not read al the replies but often see the formula you mention with a k=0.5 for natural gas.
Formula is from Hassenstein (various K factors for various fuels).
This gives you the the losses of your boiler in the exhaust stack. Your calculation give 9.7% mine somewhat less.
From that you have to deduct other losses:  radiation 0.5 to 5 % (2%)  blowdown 0 to 4 % (2%)  unburned 0.5 to 3 % (0.5%)  fouling of heat exchangers 0  4 % (1%)
Between brackets normal values for NG boiler (depends on your installation). Thus overall efficiency is roughly 85 % on LHV as mentioned by others if boiler is well maintained.
This formula is issued by several organisations to control boilers.
To go from LHV to HHV divide by 0.902.
Thus 85% / 0.902 = 94.2% on HHV.
Don't know all the theory behind the things but they work.
Hope this helps. 

25362 (Chemical) 
10 Jan 05 14:22 
With regard to the formula brought by toohotforme, I think the logic goes as follows:
1. The divisor: % CO_{2} in flues at constant xs air is directly related to the hydrogen:carbon atomic ratio of the hydrocarbon fuel. For example, for methane the ratio is 4. So, for xs air=0, the maximum CO_{2} would be 11.7%, for xs air=20%, the value would be 9.6.
The H/C ratio is directly related to the calorific value. The conversion of units is probably included in the K factor.
2. The dividend: The difference of temperatures between flue gases and air is an indication of the enthalpy delivered by the combustion of the fuel. Losses, as listed by msonnet, are probably incorporated in the K factor. 

quark (Mechanical) 
11 Jan 05 6:27 
toohot, The efficiency was based on Net(or lower) Calorific Value. Manufacturer's specification was better than(?)91% with NCV. No, I don't have the derivations for the constants. The equations are taken from databook of one of the reputed manufacturers in India(This firm has technical collaboration with Sarco). They refer to some BS code. GCV to be expressed in btu/lb. Can work for any fuel. Basically, efficiency calculation of a boiler has many caveats and it is even difficult for me as ours is a batch plant with highly variable loads. Variable condensate recovery made the problem much difficult. Further, we never accounted for wet steam and we presumed it was totally saturated. Calorific value of fuel oil was not actually tested though the density may vary from 0.89 to 0.93. Additives were being used during the efficiency test. Though I never considered the result as authentic, it gave me clues about energy leaks in the entire system. ASHRAE 2003 Handbook indicates, for a flame modulation boiler, efficiency at part loads is better due to higher heat transfer area to energy input ratio and the efficiency reduces as you go on increasing the load. Regards, 

25362 (Chemical) 
11 Jan 05 7:03 
I'm aware of the fact that european and american technicians tend to use different fuel calorific values for the definition of the efficiencies. 

This is good stuff! Thanks for the valuable information! quark's last para. about the efficiency improving at lower loads makes sense and bears out the logged data which follows this trend. msonnet's Hassenstein formula has been revealed to me by a boiler manufacturer  you are right about it. Here's more for info.  the fomula is... 100K(TGTA)/CO2% where K is the factor refered to by msonnetTG = stack temp in C TA = ambient in C CO2% = CO2 expressed as percent For natural gas the K factors are related to the CO2 as follows... CO2 K 8 0.457 9 0.466 10 0.476 11 0.486  all near enough to 0.5! The K factor seems to consider the sensible heat loss and omit to consider the H and H20 in the fuel and air. I am told by the boiler manufacturer that it is based on the LCV. The other European method is the Siegfried formula as described in one of the downloads above which takes O2 as a proxy rather than CO2 and also omits the H and H2O in the fuel and air. There is a British method that takes the above and the H and H2O in the fuel and air, giving a difference of roughly 10% vs the European methods above  which seems to correspond with msonnet's factor of 0.902 to go between LCV and HCV. I have discovered that 25362 is correct on the USA and European's working on different fuel heating values  I understand tht USA works on HCV and Europe on LCV  which seems to be borne out by the Hassenstein and Siegfried formulae not accounting for moisture loss  and thus returning high efficiency values. The way you describe the chemical influence makes it easy for a mechanic like me to understand the mystery  thanks! 



