Turbine Overpressure Protection 1

[AntMan]

Does anybody know of any methods to reduce the required relief capacity of a relief device on a turbine exhaust to a value much lower than the full flow, to protect the exhaust casing and outlet ducting in the case of a blocked outlet?

 

[poetix99]

The scenario of "blocked outlet" obstructs the full flow. The relief device must accommodate the maximum possible value of "full flow". It IS that simple.

I assume that the problem you are trying to solve is the cost of a relief valve. You could try an exendable relief device, such as a rupture disk. Over the long run an expendable (one use) device could be much more expensive than a valve.

Industry standards such as NEMA SM23 or SM24 prescribe the sizing of full flow relief valves.

If there was a cheaper safe way to provide over-pressure protection, it would be widely implemented.

 

[AntMan]

poetix99: Thanks for the tip. I believe what you have implied in your first statement is that "full flow" relief is not necessarily equal to or close to the normal operating flow. In my pursuit of the most cost effective solution, I am hoping this means a lower flow rate as the delta P across the turbine decreases, and other dynamic conditions occur when the outlet valve is closed. For example this turbine supplies power to the air compressor, which is the source of pressure. Am I also correct to conclude that the relief evaluation is more complicated than if this were a centrifugal pump?

 

[poetix99]

Hmmm...

In general, the sizing and application of a full flow relief device on a steam turbine exhaust ought to be completely independent of the driven equipment.

I am concerned that either I do not understand your application, or you need to better understand the use of relief devices. Your statement "For example this turbine supplies power to the air compressor, which is the source of pressure." is particularly confusing. Are you using compressed air for motive power of an actuator on the relief valve? This is not "fail-safe", and is not recommended (the relief valve should be passive).

Consult NEMA SM23 (mech. drive service), Section 8 for an example of the way that such requirements are spec'd. The clear INTENTION of such codes and standards is the proverbial "worst case scenario" of maximum turbine flow, consistent with conditions that pertain when the valve is "called into service".

You get to decide what is a (reasonable) worst case, within the definitions provided by the governing code, and subject to second guessing by liability lawyers. Mass flow AND steam temperature & pressure must be specified for the proper valve.

What I have seen as "customary" is that the OEM specifies flow and steam conditions for the valve to be "purchased by others" (i.e. the purchase or end user), but this is of course a commercial consideration subject to negotiation.

In my own opinion, yes, it is reasonable to recognize that, for a back-pressure turbine with a relatively small overall pressure ratio, any change in exhaust pressure will effect the flow capacity of the inlet. You should not use a back pressure that is greater than the "full open pressure" of the relief device; in fact, some margin of conservatism is advisable.

Did I miss the boat completely concerning your comments about the air compressor?

 

[AntMan]

This is a CO gas expander turbine for cat cracker regenerator flue gas. The turbine drives the air blower for the regenerator, which has a helper steam turbine. The CO gas expander exhaust includes a rupture disc upstream of the first block valve. However, for some undiscovered, undocumented reason, it is sized for only ~15% of the normal operating flow, and I am trying to recreate the original designer's relief evaluation. I know by today's standards, we would assume the exhaust valve can be inadvertently closed at any time, not just a turndown condition.

If this were a relief valve for a centrifugal pump to protect piping and downstream equipment from a high deadhead pressure, I'd look at the pump curve to find the flow at the point where the valve is relieving. This could reduce the required capacity below normal operating flow. However, in the case of a gas turbine expander, I have no knowledge/experience of whether a similar evaluation can be made, nor have I ever seen a "turbine curve," nor do I have strong hopes that it would reduce the required flow capacity as far as 15% of normal. But, we tend to assume our predecessors from several decades ago made reasonable design choices, and would like to have strong argument that they were unreasonable, at least by today's standards, before installing a much larger relief device.

What are your thoughts?

 

[poetix99]

OK, now I understand the air (or FCC gas) as "the source of pressure".

Well, FCC strings are interesting applications in general.

I am familiar with a few of these applications where the steam turbine is either a "helper" or a "starter" due to the bootstrap start necessary for (at least some) FCC strings. I understand that the gas expander is what you're interested in. So, obviously, NEMA (SM-23) doesn't apply. (We'll come back to the steam turbine, however.)

I think that it is necessary for you to understand the dynamics of your particular string. If the expander exhaust block valve is closed, the air compressor obviously isn't going to run for very long, but perhaps long enough to do some damage.

And one needs to then consider the steam turbine; is it a "helper", and running all the time, or is it a "starter" and just windmilling with cooling steam flow (minimal flow to remove heat due to blades beating dead steam)? A helper would continue to drive the compressor (at least a little).

What is the pressure rating of the expander casing?

Is there a relief device at the compressor? In the cat-cracker? Depending upon valving (including NRV's) in the cat cracker piping, these components are possibly all a part of the same "pressure vessel", to use that term loosely. Some other component beside the expander might be the limiting pressure containing component.

I'm not necessarily saying that your 15% SRV in the expander is OK. FCC applications might not be unique, but they're certainly different from applications with separate power and process loops. You have some other things to check out.

 

[negotiator]

A feed-back circuit will satisfy this. Create a smart-flow
valve to peak at any level, as-well-as a safety circuit.
Audible or visual warnings could be built around a
LM3914 and a 10-LED bar, & a few transducers. R. S. has
all the stuff, and if you need schematics, ASK.

thestocksguy@hotmail.com