10 replies to this topic

[sd_shreyas]

Dear Members,

 

A very Happy New Year to all of you. Wishing you a prosperous new year ahead.

 

I have problem related to BFW (followed by Steam) reverse flow to Deaerator and want ot protect Deaerator from over pressurization. Currently project in the design phase. I am looking for an answer from the industry people.

 

Problem:

BFW from Deaerator is supplied to various consumers like HRSG, Incinerator, Aux Boilers, De-superheaters through High pressure/energy BFW pumps. BFW supplies water to steam drums of HRSG, Incinerator and Aux Boilers. Steam drums are provided with Low Level trip but it is currently configured to cut-off the firing of burners on their activation. Steam drums RV set pressure is 28 barg and steam produced at 20 barg.

 

Each BFW pumps do have individual NRV and ESD valve in their discharge line.

 

Now, in case of BFW pump trips, their is a chance of reverse flow of BFW followed by Steam from high pressure consumers to Deaerator. This could overpressurize the deaerator. Deaerator design pressure is 6 barg (same as LP steam design pressure).

 

I feel and heard that Deaerator normally do not have the Relief valves against over-pressure. However, I want to understand the industry followed practice to protect deaerator from over-pressurization due to:

1) BFW followed by Steam reverse flow from consumers. (Plus, how should I calculate the reverse flow?)

2) High shut-off pressure of pumps feeding water to Deaerator

 

Pls share your views and related references.

 

Thank you.

Shreyas

 

[sd_shreyas]

I forgot to add that, I am not taking credit of NRV and ESD valve in discharge line of BFW pump while sizing for reverse flow.

[S.AHMAD]

Hi,

 

1. The BFW pump as you mentioned has NRV. This is the purpose of the NRV to prevent reverse flow. If the discharge pressure is above 1,000 psig, we normally install 2 NRVs. Otherwise 1 NRV is enough.

2. Are saying about BFW shut-off pressure? nothing to do with deaerator pressure.

2. For deaerator you need to consider fire case and stripping steam break-through due to wide opening of control valve.

Edited by S.AHMAD, 01 January 2014 - 12:41 AM.

[sd_shreyas]

Ahmad,

 

1. 2 NRV seems okay but still not taking credit for them.

2. I meant, Deaerator Feed pump shut-off pressure which is having high shut-off pressur ethan deaerator design pressure

2. Fire case is ruled out since Deaerator is elevated more than 25 ft. Stripping steam break-through protection not required since Deaerator design pressure is same as the design pressure of steam which has the RV upstream.

 

Regards.

[S.AHMAD]

Hi.

1. Feed pump pressure is also not relevant to deaerator over-pressure since deaerator is normally vent to atmosphere and vapor space above the liquid level. Unless you are considering the case of vent blockage and deaerator water raised up and causing overfilling (double jeopardy). How are you controlling the water level in the deaerator?

2. How do you control the deaerator pressure? are you using control valve at the vent? or an orifice plate?

3. I suggest you consider overfilling of deaerator water due to failure of BFW pump or level control valve fully open case. Otherwise you have no case of PSV requirement. so just install 3/4" x 1" PSV to meet the ASME Code requirement.

4. If you are considering reverse flow, it is also a double jeopardy case due to simultaneous pump and NRV failure

Edited by S.AHMAD, 01 January 2014 - 01:35 AM.

[fallah]

Shreyas,

 

Apply two dissimilar NRVs, but still you might have reverse flow if downstream pressure would be higher during BFW pumps trip. If the deaerator hasn't been sized to withstand downstream pressure; in sizing of the PSV on the deaerator the overpressurization due to reverse flow should be considered...

 

If you can upload a simple sketch of the system it can help more...

[S.AHMAD]

What about the LP steam. Do you generate LP steam or is it let down from HP steam? for later case there is possibility of let down valve stuck open.

[fallah]

4. If you are considering reverse flow, it is also a double jeopardy case due to simultaneous pump and NRV failure

 

AHMAD,

 

OK, simultaneous pump and NRV failures could be a double jeopardy case but check valve seat leakage should be considered. Then, check valve seat leakage can be ignored in reverse flow condition provided that the isolation valve in pump suction line to be closed in order to isolate the low pressure part upstream of the check valves from high pressure part at downstream...

[S.AHMAD]

Fallah,

 

1. Yes agreed with your comments. However, in my view, the quantity of steam to flow through a leakage is small coupled with the pressure drop across the centrifugal pump (which acts like a steam turbine). I think is going to be very small. Calculation shall be carried out to verify whether the reverse flow rate is justifiable which I think there are others credible scenarios. Looking at the P&ID should gives better. view.

2. If the NRV is the concern, installing an additional NRV should minimize the probability of occurrence.

3. Please take note that NRV is identified as a "Safeguard" in HAZOPS. If we have any doubt on any safeguard, proper way is to conduct LOPA

Edited by S.AHMAD, 01 January 2014 - 08:19 PM.

[papillion]

I agree that steam backflow to deaerator will be unlikely to occure, considering with your protection. What likely to happen is backflow of BFW from your other parallel BFW pump to deaerator via your tripped pump. This scenario once happened in our plant. Steam drum pressure 123 kg/cm2g, normal running pump tripped, stand by pump start, but no flow, since the ARCV (auto recirculation control valve) failed, and back flowed to deaerator instead of steam drum. No steam back flowed to deaerator encountered.

 

Like ahmad said, a check valve is identified as a safeguard (exception is made when you have many experience to a particular check valve application that frequently failed), than you can consider to add more safeguard.

[aroon]

Sometime, dedicated PSV may not be required in the event of check valve failure. This is because of the following couple of reasons:

1. At design pressure, steam will cease due to steam limiting pressure is same as that of design pressure. So there will not be de-aeration and credit of normal vent can be taken if leakage rate is in consideration.

2. The stripping steam is coming directly from LP steam header, which has RV (per Post #4). The credit of the same RV can be taken if there is no major restriction (control valve/RO) in between this RV and De-aerator.

 

PS: Shreyas, I guess you must have considered "Steam collapse" effect while deciding vacuum rating of the de-aerator. This just for your information.