7 replies to this topic

[chem55]

I am designing a Steam system for Batch Plant which require small Boiler (~4MMBtu/hr).
As per the design the condensate return from the Process side is going directly to Boiler Feed Water Tank (w/o the use of Flash vessel/pump package). It is a small boiler so it doesn’t have deaerator.

Condensate from Steam Traps going directly to Boiler Feed Water Tank, so Tank is also acting as a Flash vessel.

The Boiler Feed Water Tank is Atmospheric and it has vent on it. I am not sure if the Boiler Feed Water Tank has to be pressure vessel.

I read in one of the article that the Flash vessel must be Pressure vessel ((Article- See Pg 5); but I believe it is true for closed system only where flash steam is recovered. In my application I am not recovering flash steam.

Thanks in advance for your feedback.

[ankur2061]

chem55,

You have failed to mention the condensate pressure returning to the BFW, only from the condensate pressure can we know how much flash steam will be generated and lost. Not recovering flash steam is an uneconomical design, if the flash steam is a significant amount.

Another aspect of a BFW tank open to atmosphere is oxygen ingress in the BFW which can lead to oxidative corrosion in the boiler. You will shorten the life of the boiler internals coming in contact with the boiler feed water having high levels of dissolved oxygen in it.

As far as the question of the design pressure conditions of the BFW tank is considered, the tank has to be designed for the highest flash steam pressure plus the static liquid head in the tank. Also the BFW tank needs to be designed for full vacuum for reasons of condensation of flash steam. Considering this, the BFW tank does not fall into the category of a atmospheric or low pressure storage tank and automatically gets categorized as a pressure vessel.

Hope this helps.

Regards,
Ankur.

[chem55]

Ankur,

Thanks for your reseponse. I have attached the sketch for the steam/condensate system. I am considering 20 psig at the outlet of the Steam trap, and steam pressure is 110 psig. I have done two phase calculation and 20 psig at the outlet of the Steam trap would be enough to transfer the condensate to the Boiler Feed Water Tank.

The Boiler and Boiler Feed Water Tank is existing (not working at present). It is a batch process and at present client doesnt want to spend money to recover the steam. They will use oxygen scavenger to control the oxygen in water.

The Boiler Feed Water Tank is existing and it is atmospheric vessel. I was not sure if I need to replace the existing Boiler Feed Water tank to use Pressure vessel. As the vent is open to ATM - withought valve, my understanding was that Boiler Feed Water Tank doesnt need to be a Pressure Vessel.

Attached Files

•  Sketch.pdf   17.89KB   92 downloads

[Art Montemayor]

Chem55:

While I agree with your proposal that you can treat the boiler feedwater tank as an atmospheric vessel, I caution you to heed the opinions that Ankur is voicing. He particularly is concerned with the probability of your process turning this tank into a pressure vessel. Some scenarios could be:

• a steam trap failing open and allowing high pressure steam to blow-through it and into the boiler feed water tank at a rate superior to the capacity of your atmospheric vent. You haven't said how you will size this vent, so this is a possibility.
• the possibility of your vent getting plugged or shut; this could cause an over-pressure or a full vacuum situation.

I believe you can have a safe boiler feed tank operating at atmospheric pressure. BUT you must ensure that the vessel is maintained that way 100% of the time. This is the way that steam locomotive engines were operated for many decades. They had no deareators or water treatment. They operated with a once-through water & steam system. Under this type of operation you are admitting to a scope of work that accepts the small boiler as a "consumable" piece of equipment. In other words, you accept the corrosion and contamination that untreated water will have on it and you will monitor, inspect, and prepare for an eventual total replacement at the appropriate time without causing a hazard to the existing operation. You can't just fail to treat the feed water knowing what we know about the possible consequences. You must take into consideration the obvious trade-off that comes with installing a low-cost, consumable boiler operation. And this is the kind of scenario that Ankur surely is warning you about in recommending that you employ a pressure vessel instead.

[kkala]

For such a small boiler (say 2 ton/h of steam) previous posts have indicated that an atmospheric feed water tank could be an option under stated reservations. A few additional remarks, if this option is adopted.
1. Once we duty-specified a tank to receive 20 ton/h of oily condensate and reduce its oil content to about 10 ppm (mg/l). Capacity of its vent was specified 20 ton/h of steam. For another atmospheric tank, receiving 110 ton/h of non oily condensate, vent capacity was specified as 25 ton/h of steam. Highest single steam passing (e.g. through individual traps, or condensate valves) was considered, for the failure event when steam was carried out instead of condensate.
2. For (say) 50 ton/h boilers of 40 kp/cm2 g steam, condensate should be deoiled to 0.5 ppm (some boiler suppliers require 0.2 ppm) and its coloidal iron content should be reduced to 0.05 ppm maximum. You may not require such strict standards for the small boiler, but impurities content had better be followed. Allowing higher steam pressure compared to the pressure of heated liquids in exchangers, etc, helps for pure condensate (e.g. it is improbable to be oily). Oil or rusty iron in feed water can create problems.
3. Follow condensate impurities by increasing / decreasing the blow-down of the boiler. Condensate treatment (if necessary) can assumingly occur on line, on the water way from pump to boiler.
4. Oxygen scavenger can be introduced into the steam drum.

Edited by kkala, 08 October 2010 - 04:59 PM.

[chem55]

Thanks for your response.

I called couple of Vendors and all of them are saying that it is acceptable practice to have Atmospheric Boiler Feed Water tank where Condensate return from Steam traps are entering directly to ATM Boiler Feed Water tank as far as it has proper size of vent – especially for small boilers (in my case it is 4MMBtuH)

One vendor suggested that Thumb rule for vent on BFW tank is: 2.5 times the Condensate return nozzle. My condensate return nozzle is 2” so vent size should be at least 6”.

However, as Mr. Art and Ankur mentioned, I checked the hydraulic calculation for case of steam trap failure assuming:
- 3000 lb/hr steam flow (which is the maximum flow required for one of my process user),
- 15 ft pipe length,
- 3 ft head loss and
- 3 elbow
–> it gives me negative pressure at the outlet of the pipe (For Pressure/Temperature in BFW tank: I considered ATM pressure and 154.8C. Reducing the pressure of saturated steam from 120 psig to 0 psig gives me 154.8 C steam temperature).

My calculation shows that at least I need to have 14” vent on 2’ x 7’ vessel. Not sure if it is correct way of sizing the vent.

[Qalander (Chem)]

Dear

Just a minor suggestion; that you

• may have two 8" dia vents with
• one normally in service and
• other set at certain pressure to open up if the need arises?

Hope this proves helping?

[kkala]

...I checked the hydraulic calculation for case of steam trap failure assuming:- 3000 lb/hr steam flow (which is the maximum flow required for one of my process user),- 15 ft pipe length, - 3 ft head loss and - 3 elbow
–> it gives me negative pressure at the outlet of the pipe... My calculation shows that at least I need to have 14” vent on 2’ x 7’ vessel. Not sure if it is correct way of sizing the vent.

1. This is a very small vessel indeed (dia=0.61, H=2.1 m, V=0.6 m3), not a small API 650 atmospheric tank (as I initially thought of). The latter would give a residence time to condensate for deoiling and probable precipitation of suspended solids, advantages not valid in the present case. The present vessel is actually a flash tank and sizing should follow this concept. Sizing of vent looks critical, so it is better be based on calculations instead of thumb rule.
Max operating pressure for an API 650 tank is 2.5 psig; mentioned 2'x7' vessel may be probably rerated to a design pressure of ~3 Barg (it may have the appropriate thickness to stand for its own weight). It may be worth while looking into this possibility (eventually performing the hydrostatic test), which would simplify things and increase safety (but the emergency vent would not be eliminated).
2. Negative pressure at the end of pipe connected to the vessel means that a lower than 3000 lb/h steam flow can pass (calculations should be conservative), not satisfactory for consumers. Probably flow will be "arranged" in some higher pressure (or piping resized).
At any case max steam flow through steam trap should not be based on max consumption, but on the steam flow when the trap fails quite open. But in the above case this flow seems to be limited by downstream piping.