Feedwater Heater Levels and Performance

[ChaseThisLight]

Recently we have began to look at the performance of our feedwater heaters, and I have a few questions. We are a 700 Net MW plant, with 2 HP heaters, deaerator, and 4 LP Heaters. For about the past year, one of the LP Heaters was at a very low level. Once we noticed this, we changed the controller set point and raised the level. It seems like we have many valves that are leaking through and not holding level throughout the system. When the level in the FW Heater is at the Normal water level according to the original manufacturer's drawing, the DCA is higher than design, and TTD is lower than design. I've spoken to the manufacturer's of each heater (no longer in business, but the company that bought them out)and received conflicting information. One says we should adjust level to performance, and another says should always maintain normal water level. I've read an EPRI study that says it is practice to control based on performance, but uneasy after hearing from the manufacturer. Obviously we wouldn't drop the level below the drain cooler, and realize raising the level too high reduces the condensing zone. Is it standard to operate based performance? If so, do you set performance based on a load curve?

And a dumb question, when calculating TTD I should use absolute pressure, right? For instance, the extraction pressure for one of the LP Heaters is 5 psia, which is what I would go the steam table for saturation temperature, but our performance software uses 5 psi + atmospheric pressure. I don't understand why they would add atmospheric, the heater is not open to atmosphere. But using just psia gives a TTD of -50 degrees F. I dont think that is even possible.

Any help would be greatly appreciated. Thank you very much!

Chase

 

[stgrme]

I do not have the proper background to comment on how to operate the feedwater heaters. However, I can make a suggestion on the use of your performance software. From your description, it appears that the performance software is set up to accept gauge pressures. It may also have an input of the local atmospheric pressure. This set-up may be the most appropriate if all the inputs to the software are gauge pressures.

If I have interpreted the set-up correctly, you must input the gauge pressure rather than the absolute pressure to the software. For an absolute pressure less than atmospheric pressure, the gauge pressure will be a negative number. (Absolute pressure minus atmospheric pressure equals gauge pressure, a negative number for an absolute pressure less than atmospheric pressure.)

Best of luck!

 

[JDSouthTexas]

Our units are 1425 MWe gross (per unit) with best Condenser vacuum.

If you ran a long time (1 year) with low liquid level steam passing through the drain cooler may have damaged it (vibration / flashing).

I would recommend controlling level based on performance...usual #s are 10 deg F DCA, 5 deg F TTD.

Look at Electric Power Research Institute (EPRI), Heat Exchanger Institute (HEI), and American Society Mechanical Engineers Turbine Damage Prevention Fossil/Nuclear(ASME TDP-1or2) references.

Too low level & you could damage FWH, too high level & you can damage your turbine (see references above).

Powerfect / Stan Yokell (MGT) do training / consulting on this topic. as do I for $ lol!

raising liquid level to prolong life of FWHs if damaged is small efficiency loss (less heat transfer area) vs. continued damage if too low level / steam enter drain cooler through existing holes...

if you eddy current / etc. FWH tubes you can see what tubes have damage / where drain cooler is damaged...

what area are you in?

 

[EdStainless]

You need to use absolute pressure throughout, some of your LPs are running in vacuum.
If you have leakage and such then are you really sure of your level control? My hunch is that you may not have been running as low as you think.

There is a tolerance on level, most plants will shade up or down a bit if it helps performance.
Make sure to check this at various power levels, and keep checking against your model. If at some place readings don't match the model you need to hunt down the defective gage or controller.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[davefitz]

An error in the level Xmitter may occur if the htr is at vacuum and air is leaking into one of the level taps.

JDSouth is correct on all counts.

The saturated liquid that enters the drain cooler is admitted through a "snorkel" and needs a few inches of gravity head to avoid flashing at the drain cooler entrance, especially if the drain level control valve is unsteady and hunting instead of holding a steady flowrate.

"Whom the gods would destroy, they first make mad "

 

[georgeverghese]

Another potential problem with level dp cells is when they are used in condensing vapor service ( i.e steam) - diaphragm seals should used on the impulse lines to avoid condensation in the vapor leg / vapor traps in the liquid leg.