Deaerator Vent Steam Recovery 1

[owg]

We are evaluating a proposal to recover most of the small amount of steam that vents with the non condensible gases from a deaerator. The feed is about 50% preheated city water and 50% recovered condensate. The deaerator runs at about 5 psig and there does not seem to be much steam coming out of the vent. The proposed recovery device contacts some of the feed water with the vent gases to recover the steam. The economics assume that 1% of feedwater is currently lost as steam. How could we accurately calculate this current steam loss. We have seen models that assume 0.5% for this loss. If this is our case it would half the proposed savings.

 

[TD2K]

How is your vent rate controlled? Typically, either a valve is just cracked open or an orifice is used. If it's an orifice, it's fairly simple to calculate.

If you have a valve, you might want to look at putting a cap on the outlet with a small hole and open the valve fully. Keep increasing the orifice till you have the flow of steam you want and then go back and calculate the steam vent rate based on the orifice size.

I'm not sure if the flow rate of the steam would be enough to justify it. Plus, since the deareator is removing dissolved gases, your exchanger will also need to vent these off the steam side to prevent gas blanketting which will further reduce the amount of heat recovered.

 

[TBP]

There's usually not as much steam to recover as it may appear. Make sure the amount of steam your venting is no more than necessary. Lots of projects like this are based on incorrect initial settings, which are fixed (often during the project), which then renders the project a waste.

The most common method I've seen for controlling DA vent steam is to drill the gate of a valve (perhaps 1/8&quot, then leave it closed. Put a valve with an un-drilled gate ahead of it to permit any on-the-run changes. If this is what you have, or can install, then publications like Spirax Sarco's "Hook-Ups" have tables that show steam flow from various sized orifices at a variety of differential pressures.

 

[owg]

I think the vent is controlled by a cracked open valve. We will check with the client. The proposed device is not an exchanger, it is a direct contact device, patented so no more details available.

How do we establish that a 1/8 inch orifice is the correct size. Don't we need to know how much steam is to be vented to do the job. How can we estimate this? Do we need to experiment?

Thanks for the comments so far.

 

[TD2K]

You are just providing a small slipstream of steam to purge the incondensibles that are stripped out in the deaerator. If you are concerned, check the outlet oxygen concentration from the dearator, it should be in the low ppb range.

Someone doesn't make sense to me. You are running feedwater through the deareator to remove oxygen and C02 and it sounds like you are then going to take that stripping steam, with those gases and do a direct injection back into the feed water to the dearator. Sounds like the gases would just build up and up in the recyle loop. Or is this feedwater stream going to another location in the plant where these gases are acceptable?

 

[owg]

Think of a small packed tower on the deaerator vent. A small slip stream of the deaerator feed water runs down through the small tower and into the deaerator. The deaerator vent gas runs countercurrent up throught the small tower and out to atmosphere. The intent is to recover most of the previously vented stripping steam as boiler feed water.

 

[TBP]

If the device is already patented, why should the manufacturer worry? Why the secrecy? I agree with TDK2 - something doesn't add up here.

Is this "patented device" significantly cheaper than a small stainless steel heat exchanger? I KNOW that will recover the heat of the vent steam, especially with cold make-up water on the other side. No "propietary device/technology" fooling around, either.

Talk to the DA manufacturer about what kind of orifice size they recommend with your model of DA, and operating conditions. If you can't get any worthwhile information, I'd try just installing an isolation valve ahead of the gate valve, and try an 1/8", and see what you get. A little trail & error should tell you what you need to know. You'll need to monitor things like operating load, and sulphite (I'm assuming you're using sulphite as an oxygen scavenger) consumption and residuals.

With a delta-P of 5 PSI (DA operating pressure to atmosphere), an 1/8" orifice will pass 7.5 lbs/hr, according to Spirax Sarco's "Hook-Ups". A 1/2" orifice will only pass 119 lbs per hour. I don't think your project will have to cost very much before the payback drifts away. I think you'll get a lot more bang for your buck just repairing/replacing a couple of failed steam traps in the plant.

 

[owg]

The manufacturer describes the device as patented, but chooses not to provide details on their website at

http://winwellglobal.com/.

I do not no why he/she would worry. I agree there are better uses for money around the plant. Thanks for all your suggestions. If I am able to get any hard data I will post it.

 

[TD2K]

Hmm, I don't have a lot of reference books but that 1% vent rate seems high to me.

We sure didn't have that from our packaged boilers IMO. They were about 500,000 lb/hr max capacity and the dearator was more than that as it also feed all the other steam generators on-site. All I could see you needing to vent is enough steam that the oxygen concentration is kept low in the vent stream. With a few ppm, maybe, in the feed water, that doesn't have to be much steam flow I would think.

Any other comments or am I all wet (no pun intended) on this one?

 

[TBP]

If you assume that a given steam plant has a 50,000 lb/hr average load, then with a 5% blow down rate, you'd have to feed 52,500 lb/hr water to the boiler. If it's assumed that 1% of 52,500 is lost as DA vent steam, then that's 525 lb/hr. That's something like a 7/16" orifice, and carrying a pressure of 100 PSIG in the DA. Unlikely.

Something that haunts calculations like this is that very few plants have any sort of accurate metering for steam. Most just have orifice plates (and fully half of those will have the plates installed backwards...) with no pressure compensation. Steam production & usage numbers in most plants are not much more than fiction.

 

[jproj]

owg:

99.95% of the units I design have a 0.05% vent rate, which is more than adequate (in most cases) to vent the non-condensable gases and reach an outlet oxygen concentration of 0.005 cc/L and ~0.0 cc/L CO2. Exceptions come with every job, but unless there is a low delta T from the inlet to the outlet, or there is unusually high non-condensable gas content, a 0.05% vent rate is standard.

Regarding the vent orifice.... the hole is drilled into the gate to insure that the vent can never be closed. In normal operation, the gate is opened to release steam & non-condensables. The amount the gate is open is generally a trial and error proceedure that involves testing the outlet concentration of non-condesable gases. Vent rates can be calculated using an orifice equation. The most common size on a DA vent valve is 2".

Regarding the "vent condenser". External vent condensers are requested sometimes and they are always specified to be shell and tube heat exchangers. They are very effective and are also not very expensive. I would shop around if I were you... especially if the vendor is not willing to disclose how his "black box" "vent condenser" functions.

You may want to make sure the vent condenser is required. Nearly every DA in service has an internal direct contact vent condenser. Besides, the steam loss is not considerably large.

Example:
A 500,000 lb/hr (outlet) DA with 50-50 makeup (60ºF) - condensate (180ºF) inlet mixture only needs to vent 283lb/hr, which uses a 4" gate valve.

Good Luck!

jproj

 

[jproj]

my apologies... I meant 0.5% vent rate.

jproj

 

[owg]

Thanks TD2P, TBP, and jproj. We got a cross check on their boiler steam make to about 2% by using their boiler efficiency readings. They were basing the readings on 1000 BTU/SCF HHV natural gas. We found that the local utility is regulated to supply at least 965 BTU/SCF and it is my bet that they come very close. Its a good business selling nitrogen at natural gas value.

The idea of an existing built-in condenser is worth a check. The 0.5% par for steam vented confirms our finding in the literature.

Thanks for the quality help. I will post when we have more information. This issue is part of a larger study.

 

[TBP]

What kind of steam meters does the plant have? Are they pressure compensated? Most plants have orifice plates without pressure compensation, and in earlier efforts to increase efficiency, many reduced the boiler operating pressure. Most of them forgot to do anything with the steam meters. If the orifice plates weren't changed, and/or there's no pressure compensation, the meter will read high.

Be very wary of the steam numbers you're being supplied with. The error can be significant.

 

[spector]

If you are experiencing a 50% make up rate, there is lots more money to be saved by reducing that rather than trying to squeeze a little water from your DFT vent stack.

Chase down your condensate leaks. Install returns where there are none and fix all faulty steam traps. Reducing your make up will also reduce your blow requirements. Why worry about what little is going up the stack when tons more is going down the drain?

 

[owg]

Orifice is pressure and temperature compensated. We agree there is flash steam to be recovered.