Gradfathering with 3% inlet drop?

[Bourbon103]

Stab in the dark:

A client has a 3-way safety selector valve with (2) relief valves, circa 1960s, on a 75 psig steam line. They want to replace the 3-way and eventually the relief valves (down the road). The failure scenario is a failure of the 300 psig/75 psig pressure reducing station OR the bypass valve full-open/fail.

The problem is the distance the relief valves are from the station causes a > 3% inlet pressure drop.

I think the owner believes that if we just change out the 3-way, and not the relief valves, that the old relief valves would be fine until they were changed out (being "grandfathered", if that's even applicable...I'm assuming the 3% rule was adopted after it was installed). Again just a shot in the dark. Safety is always most important. What do you guys think?

 

[Latexman]

The 3% rule has been around a long time. It is closer to Good Engineering Practice than it is mandatory by Code. Some companies impose their own interpretations on this though, mine included. Knowing the calculated % following the installion of the new safety selector valve would help us color our advise the correct shade of gray though.

Good luck,
Latexman

 

[pleckner]

Shade of gray it may be but I have been doing process relief analysis and design for almost 20 years now and I have never read anything that allows "grandfathering in" when it concerns safety. I would request the client to tell you in writing that you do not need to follow the 3% Rule. Yes, it is not mandatory but is as Latexman wrote considered good engineering pracitce. And yes, it isn't absolute but that only means you can have higher losses as long as you can prove you have not created a situation where the relief device won't function properly, e.g the blowdown of your particular valves are high enough not to be affected by the larger losses.

 

[4Pipes]

Just to add to the very good replies. The only bit I would comment on is calling it good engineering practice. As Pleckner said, it has a solid basis of fact rather than a suggestion that it is preferable. That said, its only a form of words and it is definitely good practice. I mention it only because several years ago, I was on a project where the 3% was regarded as preference that was only met if layout allowed.

Another angle on the original question, is the old chestnut of what flowrate. Everybody has heard of the 3%. But at what flowrate? The flowrate for the 3% is clearly specified in the codes. However, my experience is that it is not always applied as strictly as the "3%" especially when the calculation is done prior to purchase of the valves.

Was the original "3%" of the correct pressure drop? If not the existing system is not correct and potentially unsafe. It might never have operated in anger. Applying grand father, test of time or whatever, might be risky.

The other variable is where is the pressure drop calculated from? Again, my experience is that this is also often not very consistent espcially down stream of equipment where the choke point is not on the outlet flange.

 

[pleckner]

I'm not sure where @4Pipes was going with his statment, "The other variable is where is the pressure drop calculated from? Again, my experience is that this is also often not very consistent espcially down stream of equipment where the choke point is not on the outlet flange."

The 3% Rule only applies to the inlet line and there is no issue with choked flow. The pressure drop is calculated from the vessel under protection to the inlet flange of the PSV and this includes the inlet loss to the vessel outlet nozzle.

And another point brought up by @4Pipes is "... especially when the calculation is done prior to purchase of the valves."

My PSV calculations are not complete until I perform checks on both the inlet and outlet lines. Yes, these are often not designed yet but as the Process Design Engineer I specify the line sizes in both cases and I state on the P&ID the maximum allowable line lengths allowed and my assumption of fittings used so the Piping engineers have something to work with. This goes a long way in preventing what may be extensive and expensive rework during detail design. For example, on one project I had 25 valves to size during basic engineering. During detailed design, AFTER piping design was completed, I had NO rework in my sysem.

Moral of the story, check the inlet and outlet lines at the same time you size the PSVs and before going out for purchase.

 

[lizking]

Not sure where 4pipes was going either; but there is always confusion when the RV is downstream of a reducing station. I think losses are always taken from the mainline to the RV (normally non-flowing inlet pipe). If RV is protecting the piping that it's mounted on, analysis should be done to determine pressures at the outlet of the reducing station for overpressure compliance. Also, because the actual relieving scenario will have normal flows continuing past the RV, it may encounter instability due to lack of sufficient flow to keep the RV open. Sometimes the use of a modulating RV is warranted. "consult the manufacturer"

 

[Latexman]

The inlet loss REALLY gets interesting on a half-pipe jacket, which is usually a separate Code vessel according to the nameplate, in a fire case!

Good luck,
Latexman

 

[4Pipes]

Sorry for not spelling it out greater detai.

a) 3% - agreed but I would consider it a technical requirement and not just a preferance or good practice. (Folk are very good at reading preferences and practices to suit their own requirements hence my caution).

b) The "pressure drop" we are comparing the 3% with is as important as the "3%". What pressure drop?
The "pressure drop" requires 2 things:
i) A flow rate
ii) A flow resistance.

IMHO the flow rate should be as specified by the code.
The flow resistance depends on where you start and end. The "end" is obviously the valve. Where do you start?
From a vessel is easy. From anywhere else it should be the choke point that limits the flow or at least IMHO. What about equipment where the "flow limiter" is not at the outlet flange. Do you start counting from the outlet flange or do you include the pressure drop between the actual choke point and the outlet flange. Half pipes as Latexman pointed out are interesting. So are turbines.

All I am saying it that in my 35 years, 3% has been consistently talked about but I have come across pretty any combination of flow rate and flow resistance you can think of. Being a simple sole, I would think it nonsense to require 3% if you don't specify what you are comparing it with.

correct (=3% or as required to suit blowdown) x nonsense = nonsense

Going back to the original post - Did the original design use the same flowrate and flow resistance ? Was it correct ?

In my experience, this is an area I would be checking to find why something was "accepted" 30 years ago but not now. It might never have been correct but never operated at its full design. Even if it has operated, it might have just chattered a bit, caused a bit of seat dampage, followed by a bit of cursing and then forgotten about.

You would only know there was a real problem if the system was not able to handle the higher upstream pressure. Often times, there is and we get away with it. Occasionally not.

 

[lizking]

The 2 criteria for calculation are "rated flow of the relief valve" and "set pressure." Also, I have worked for companies that allow 7% or the % blowdown, whichever is lower, as the criteria for "existing" RV's. Do you call this "grandfathering"?