7 replies to this topic

[Saml]

This is the application:

 

A steam header above 100 bar pressure is feeding compressor drivers.

 

When one of these turbines/compressor trips, the pressure in the header goes up quickly and since both the steam generator and the pneumatic pressure control valve are slow to respond, it is usual to get to the PSV set point. These are ASME I valves without blocks. So if they do not re-seat, it requires a total shutdown of the unit.

 

The main suggestion is to use a quick acting vent (an hydraulic one) instead of the pneumatic valve.

 

However an option that has been suggested and has met a lot of questioning is using an additional relief valve set below the MAWP that has the function of a control device (in addition of the valves required by code), supplied with adequate blocks so it can be maintained if it does not re-seat properly.

 

Apart from the difficulty of obtaining reliable block valves to isolate 100+ bar steam that allow the safe removal of the relief valve. Do you see anything fundamentally wrong with this approach?

 

[Bobby Strain]

You could vent the high pressure steam. But if the turbine has extraction and/or condensing, you should consider what happens to those headers. For the correct solution, you need a dynamic model for your steam system. And that is not cheap or easy.

 

Bobby

[Art Montemayor]

By proposing to use a safety relief device as a control instrument you are misapplying the basic principle and basis for safeguarding a continuous process operation.

 

Safety relief valves are not - and have never been - designed or intended to be used as control devices.  They are meant as emergency devices and if a total process shut down is provoked by their action, then so be it.  By installing a PSV to function as a pressure relief when your continuous operation needs it is to impose a control action on a safety device.   I strongly warn you not to use that sort of twisted logic.  There is a time and a place for a safety device and there exist control devices that are designed and meant to control a process when operating conditions vary and need control.  Your described desired operation calls for process control of an expected pressure deviation and one way to counter this effect is as Bobby Strain states: vent the excess high pressure steam to the atmosphere.  However, I would venture to state here that Bobby recommends is that you do this with conventional control devices - and not safety valves.

 

I detect an experienced safety comment in Bobby's recommendation to seek the appropriate solution in a dynamic model of your operation and I concur with his thoughts on your query.

[farid.k]

This is the application:

A steam header above 100 bar pressure is feeding compressor drivers.

When one of these turbines/compressor trips, the pressure in the header goes up quickly and since both the steam generator and the pneumatic pressure control valve are slow to respond, it is usual to get to the PSV set point. These are ASME I valves without blocks. So if they do not re-seat, it requires a total shutdown of the unit.

The main suggestion is to use a quick acting vent (an hydraulic one) instead of the pneumatic valve.

However an option that has been suggested and has met a lot of questioning is using an additional relief valve set below the MAWP that has the function of a control device (in addition of the valves required by code), supplied with adequate blocks so it can be maintained if it does not re-seat properly.

Apart from the difficulty of obtaining reliable block valves to isolate 100+ bar steam that allow the safe removal of the relief valve. Do you see anything fundamentally wrong with this approach?

Is it quick opening control valve already applied? For me, slow and fast response can be controlled by instrument. Maybe can add some pressure transmitter @ different locations along the header and make voting.

[Saml]

Bobby, Art.

Thank you very much for you insight.

 

Bobby: there are hydraulic let down stations that tracks the steam usage and extraction of the turbines, have been characterized and take action to maintain the pressure of the intermediate pressure header at 45+ bar. That part is already covered.

 

The main issue here is that when the PSV pops at the steam generation drum and do not re-seat, a total shutdown is needed. Today the setup is to vent the excess steam to atmosphere by a control valve. However, it has been long discussed how fast can that control valve needs to be to avoid the main relief valves from acting.  All the discussion lead to an hydraulic operated system with high speed signal polling. We have gone thru many discussion about anticipating action triggered from the compressor trip, changing the actuator size, installing an air reservoir near the valve, etc, etc.

 

Art:

 

Agree that for a control function, you need a device designed for that fuction. So let me rephrase.

 

The purpose of the the vent valve is to remove the excess steam to avoid the actuation of the relief devices on the steam generator. That is, keep the pressure below MAWP.

 

If those valves were ASME VIII devices we would be looking into the exceptions of appendix M to maintain the valves whithout shutting down the unit. This is not possible on ASME I devices.  We need a quick acting device to prevent pressure for rising or tolerate the consequences of the PSV popping.

 

We have received the  proposal of of using a relief device for this function with a good deal of questioning and doubts.  However there are people arguing that the ASME code does not forbids the usage of a device below the MAWP, nor set any restriction on those (like prohibiting block valves) as long as the main relief valves are rated for the full relief requirements and comply with all the code mandates. 

The code, the argument goes, does not dictates how to control the pressure below the MAWP. It only states that when the pressure reachs the MAWP there should be relief valves with a set of requirements that must open.

In a way, the PSV with a lower set point would be a "control" device and not subjected to the code requirements.

 

The consequences, the argument keeps going, would not be worse that the current opening of the main PSVs. The alternative is installing an hydraulic valve on a elevated point or bringing the pipig to grade.

 

My gut feeling is that this is not a good solution, I have not seen this arrangment anywhere before, but I cannot find a good argument to say "no, because we are violating (whatever) we need to go with the more expensive option".

[Bobby Strain]

Voting is not good engineering practice. What you need is good information. Sounds like nobody in your organization is willing to spend the time and money necessary to solve the annoyance. And, never trust your gut. Most likely that feeling comes from something recently eaten. So now you have suggestions from 2 people with combined experience of more than100 years. You don't often get that.

 

Bobby

[farid.k]

Voting is not good engineering practice. What you need is good information. Sounds like nobody in your organization is willing to spend the time and money necessary to solve the annoyance. And, never trust your gut. Most likely that feeling comes from something recently eaten. So now you have suggestions from 2 people with combined experience of more than100 years. You don't often get that.

Bobby

Noted sir..may i know why it is not good practise?

But i trust, problem above more or less just like surging at discharge compressor. Engineering wise for protection is by providing instrumental for detection of the pressure and open the appropriate control valve to avoid surging.

[Saml]

And, never trust your gut. Most likely that feeling comes from something recently eaten.

 

It is always funny to read your harsh humor. Even when this time it is directed at me. While it is difficult to convey how you feel in a forum, believe me:  you made me laugh this morning.

 

Thank you very much again.