Real time boiler efficiency using PTC method?

[seasar]

The "powers that be" monitor our boiler efficiency and our current method uses the natural gas and steam flow meters. Due to very dramatic swings in steam load, we often have a boiler firing in it's low range (20% or below). The steam flow meters simply don't work at that range and I get calls asking why the boiler was 5% efficient yesterday. Explanations do very little good (just have your tech calibrate it better).

Given that I have real time NG flow from reasonably reliable meters (they match the gas bill within a couple percent), stack gas temp and excess O2 I should be able to use the stack loss method (ASME PTC 4.1). I can't find this on their website and what I find on the web implies that I have sampled my flue gas for it's constituents and their fractions. Since this is natural gas and I even have the sampling reports for it shouldn't I be able to calculate boiler efficiency online in the PLC? If so could someone help walk me through the process? I fully intend to average any readings over the span of a day knowing that would add a bit more validity. I'll also adjust this down using the rule of thumb number provided for boilers for radiative and convective losses.

Thanks for any help...I'm pulling my hair out trying to find something to go by.

 

[athomas236]

It would be useful if you could tell us a little about your boiler. Information such as the following would be useful:

1. Boiler outlet steam conditions of pressure, temperature and flow rate.
2. Boiler feed water conditions of pressure and temperature, assuming at this stage that the steam flow is approximately equal to the feed water flow.
3. Does the boiler have an airheater and if it does, is it of the rotary or tubular type.
4. Where is the flue gas temperature measured and is this the same location as the oxygen measurement.

It would also be useful to know your objectives for on-line efficiency measurements. For example, are you trying to reconcile the actual boiler efficiency with the measured natural gas consumption or are you trying to monitor any deterioration in boiler efficiency over the long term.

For large industrial natural gas fired boilers, the PTC4.1 efficiency is typically 85% of which the largest losses are the wet and dry gas losses at about 10% and 4% respectively. To calculate these losses requires analysis of the natural gas (and hence the GCV), and the flue gas temperature and O2 content in the flue gas. In the absence of the analysis of the natural gas, the GCV and combustion air requirements could be inferred from on-line measurements of specific gravity or the results of your sampling could be used to determine these.

Best Regards,

athomas236

 

[davefitz]

There are a several of sources of large errors in using standard plant monitoring equipment to determine realtime boiler efficiency.

First, steam flow measurement is not as accurate as using feedwater flow measurement minus blowdown. If you insist on steam flowmeasurement, ensure the flow meater is pressure and temp corrected and the ISA calibration curve for the flow element is programmed into the DCS.

Second, fluegas thermocouples have a huge error- first due to corrosion of the bimetallic element, second due to gas temperature stratification- multiple sampling thermocouples may be needed to lower the error. The error is very large at the outlet of a rotary air heater.

Errors can also occur with O2 meters - of course, dry vs wet correction must be used. But there can also be O2 stratification if there are multiple burners or multiple sources of overfire air- so good mixing upstream of the O2 meter must be ensured if a single O2 meter is used. There are now availble gas mixers , originally designed to assure good ammonia mixing upstream of an SCR , but can also be used to ensure accurate O2 and temp measurements- See Evoniks/ Steag products.

Don't forget that as load drops, the radiation loss term Lr increases to very high values- so this curve ( in ASME PTC 4.1) must be programmed into the realtime DCS calc. If any money hangs on this calc, then it is a good idea to perform an IR survey at full load to confirm the true radiation loss from this unit- very important if many hot hopper are used, as with CFB boilers.

But the error described in the intial post suggests the flow elements are not properly corrected for pressure or temperature.

 

[seasar]

AThomas236 and davefitz,
1. Boiler outlet steam conditions of pressure, temperature and flow rate. Saturated 150 psig, flow between 10,000 and 80,000 lbs/hr
2. Boiler feed water conditions of pressure and temperature, assuming at this stage that the steam flow is approximately equal to the feed water flow. P ~ 240 psig, T ~ 240F
3. Does the boiler have an airheater and if it does, is it of the rotary or tubular type. No air heater
4. Where is the flue gas temperature measured and is this the same location as the oxygen measurement. Post economizer, ~ 300F (a second element is installed pre-economizer) the O2 is measured pre-economizer. I believe any stratification would be eliminated after going through the economizer

The steam flow meters are P & T compensated and appear to work fine when the boiler load exceeds 25-30% (they match the feedwater flow meters)...the issue occurs when boiler load drops below 20%.

Thanks for your help,