12 replies to this topic

[kahlilj]

Can anyone explain (or direct me to a source that will) which valve characteristics should be used to determine a relief valve's capacity - API orifice size & discharge coefficient OR the "Actual" orifice size & discharge coefficient?

i know that typically manufacturers will express their product's capacity using API numbers, but in the user's application is it necessary to go with API or can "Actual" be used?

any help is greatly appreciated.
kahlil

[Azam]

The PSV vendors generally use effective areas in their sizing calculations and thus apply a effective coefficient of discharge (K~0.975) which are consistent with API recommended practice 520. If you have in your calculations used the K factor and other capacity correction factors like due to back pressure or steam (if its steam) than the area you calculate is alright for all practical purposes.

SAA

[Azam]

Kgj

1) API developed a series of inlet, orifice, outlet combinations for various flanged valve pressure classes which are utilized throughout the petroleum and hydrocarbon industry. These standard sizes are characterized by a series of fourteen standard letter orifices ranging from D through T. As an example the effective area of a J orifice is 1.287 sq. inch. This orifice area is used in API formulations to calculate valve flow rate. The manufacturer is not required to produce a valve with bore area equal to effective area. Rather he is obliged to produce a valve which will have a flow rate equal to or greater than that determined by the API formulation.

2) API formulations use Kd=0.975 for gas & vapour service, for liquid service they use Kd=0.62 however, as said in point1 the effective areas and API assumed Kds are generally different from actual orifice areas and discharge coefficients that are used to determine certified valve capacities. Effective areas calculated by use of API formualtions and its Kds will always result in selection of valves with certified capacities equal to or greater than required capacities. The effective-area concept in API allows for selection of valve size independent of manufacturer.

Regards.
SAA

[pleckner]

I would get out of the practice of using API effective orifices and the corresponding effective coefficient of discharge. All the relief valave vendors are now moving towards specifying their valves in terms of the ASME or actual orifice areas with the corresponding certified coefficient of discharge (which is nothing more than 0.9 times the effective coeficient of discharge).

The equation to calculate area is exactly the same. The only difference will be is that if you use the certified coefficient in the equation, your answer will be the ASME orifice area. You would then choose the actual orifice based on that manufacturer's table for his ASME orifice sizes.

Let me give you a real world situation involving a Farris valve. I can't remember the exact numbers but we calculated a required orifice of somewhere around 0.13 sq. in. The D orifice is 0.11 sq. in. so we had to go to an E orifice (0.196). But when the vendor came back, he said we can still use a D orifice. You see, the Farris ASME D is 0.15! Yes, we calculated the 0.13 using the Farris coefficient of discharge (0.973). But to convert this to ASME, you just multiply 0.973 by 0.9 to give you a certified coefficient of 0.876. The calculated ASME area became something like 0.144 sq. in. Still good for a Farris D orifice.

Would it have been more conservative to choose the E valve? Yes but it was not correct to do so. Also, oversizing relief valves is not a good thing.

The other advantage of staying with ASME orifices is that once you choose the standard orifice (the next largest size), all you need to do to get what will be the stamped capacity is ratio the two areas to the required relieving flow. People used to do this all the time with the API orifice but they were wrong in many an instances. The ratio only works if the API orifce/ASME orifice = 0.9 and it just isn't always so.

[kahlilj]

sorry i didn't reply sooner, but thanks Phil & SAA for your advice. it has been helpful.

kgj

[Bruce]

Having just been through this exercise recently, the answer to when to use API or ASME values is clearly defined in API-520, Pt. 1, sec 3.2. Paraphrasing, the API values are "effective" values used for initial selection. They do not represent any specific real world valve but have been generated to represent the typical PSV. ASME values, on the other hand, have been developed through certified flow testing.

API clearly states that the actual (ASME) orifice area and rated coefficient MUST be used to verify actual capacity. Effective (API) and Rated (ASME) values should NEVER be mixed in the sizing equation.

In other words, use API for preliminary estimating but confirm the purchase with ASME.

The confusing issue is that, in the sizing equation, API uses Kd while ASME uses K. The 0.9 difference (K = Kd x 0.9) apparently is rolled into the API effective area, which is roughly 10% to 15% smaller than ASME actual orifice area.

Comment to Pleckner about the Farris D-orifice getting you out of a jam. This particularly valve appears to be an anomoly. Maximum capacities of nearly all the other letter-orifice valves (Crosby, Farris, and Consolidated) are within several percentage points of API estimated capacity.

Hope this clarifies the issue.

[pleckner]

I'm sorry but there is no longer any reason to use the API values just to get a preliminary size. One can just use the ASME coefficient of discharge and get the ASME area directly. The vendors are now moving toward this in their cataloges. You will find both the API effective areas (the letters we have all come to know and love) and the ASME areas published so you can choose the correct valve the first time.

And, it isn't just the Farris "D" orifice that has this problem. I admit it gets better as the valve gets larger but there are still problems. And I remember checking Crosby some time back and it didn't work for them either. And besides, you did say "...several percentage points.." Not good enough for me. I want to use the ASME certified capacity in my calculations, not those within several percentage points.

[Bruce]

Pleckner,

I assume that you work directly for a company that uses a specific PSV vendor so that using ASME rather than API values makes sense. I work for an engineering company and the vendor often is not identified until PSVs have been bid and selected. Thus, we use API for sizing estimates (if we were to use ASME values, whose would we use before the manufacturer has been selected - that's the purpose of API providing a set of K's and A's) After vendor selection, we work with the rep to confirm sizing. Unless the client preselects the PSV vendor so that we can use ASME values from the get-go, we will continue to use API.

Apparently I was not as clear as I thought with my comments on the Farris D-orifice. For a given relieving rate and condition, the sizing equation reduces to A x K = constant. Assuming maximum PSV flow, the API estimate and ASME actual can be compared by ratioing API AxK and ASME AxK. Using Red Book values for Farris, Crosby and Consolidated, I found that the ASME/API ratio ranged from 0.97 to 1.06 with most of the values falling in the .98 to 1.02 range. The only anomoly was the Farris D-orifice with a ratio of 1.2. This tells me that API values are fairly representative of real world valves for estimating purposes but any valves sized near maximum capacity by API could change size under ASME confirmatory sizing.

The "... several percentage points" was simply saying that API and ASME sizing are close but, of course, final sizing MUST be confirmed with actual ASME values.

Bruce

[pleckner]

We're going to be pretty much in agreement on this one. I still have issues with the ASME/API ratio but that's for another time when I have the time. I want to review the API orifice sizes for the "others" before I comment further.

By the way, I've been working in E&Cs for 24 years! 99.9% of the clients I've done projects for have preferred vendors. And it has only been relatively recent that I've begun to use the ASME orifice in selecting a PSV. If there is a choice, then you are definitely correct, you'll never know until the low bidder is selected.

[fallah]

I would get out of the practice of using API effective orifices and the corresponding effective coefficient of discharge. All the relief valave vendors are now moving towards specifying their valves in terms of the ASME or actual orifice areas with the corresponding certified coefficient of discharge (which is nothing more than 0.9 times the effective coeficient of discharge).

The equation to calculate area is exactly the same. The only difference will be is that if you use the certified coefficient in the equation, your answer will be the ASME orifice area. You would then choose the actual orifice based on that manufacturer's table for his ASME orifice sizes.

Let me give you a real world situation involving a Farris valve. I can't remember the exact numbers but we calculated a required orifice of somewhere around 0.13 sq. in. The D orifice is 0.11 sq. in. so we had to go to an E orifice (0.196). But when the vendor came back, he said we can still use a D orifice. You see, the Farris ASME D is 0.15! Yes, we calculated the 0.13 using the Farris coefficient of discharge (0.973). But to convert this to ASME, you just multiply 0.973 by 0.9 to give you a certified coefficient of 0.876. The calculated ASME area became something like 0.144 sq. in. Still good for a Farris D orifice.

Would it have been more conservative to choose the E valve? Yes but it was not correct to do so. Also, oversizing relief valves is not a good thing.

The other advantage of staying with ASME orifices is that once you choose the standard orifice (the next largest size), all you need to do to get what will be the stamped capacity is ratio the two areas to the required relieving flow. People used to do this all the time with the API orifice but they were wrong in many an instances. The ratio only works if the API orifce/ASME orifice = 0.9 and it just isn't always so.

I think you mean ASME orifice/API orifice=0.9

[CMA010]

I would get out of the practice of using API effective orifices and the corresponding effective coefficient of discharge. All the relief valave vendors are now moving towards specifying their valves in terms of the ASME or actual orifice areas with the corresponding certified coefficient of discharge (which is nothing more than 0.9 times the effective coeficient of discharge).

The equation to calculate area is exactly the same. The only difference will be is that if you use the certified coefficient in the equation, your answer will be the ASME orifice area. You would then choose the actual orifice based on that manufacturer's table for his ASME orifice sizes.

Let me give you a real world situation involving a Farris valve. I can't remember the exact numbers but we calculated a required orifice of somewhere around 0.13 sq. in. The D orifice is 0.11 sq. in. so we had to go to an E orifice (0.196). But when the vendor came back, he said we can still use a D orifice. You see, the Farris ASME D is 0.15! Yes, we calculated the 0.13 using the Farris coefficient of discharge (0.973). But to convert this to ASME, you just multiply 0.973 by 0.9 to give you a certified coefficient of 0.876. The calculated ASME area became something like 0.144 sq. in. Still good for a Farris D orifice.

Would it have been more conservative to choose the E valve? Yes but it was not correct to do so. Also, oversizing relief valves is not a good thing.

The other advantage of staying with ASME orifices is that once you choose the standard orifice (the next largest size), all you need to do to get what will be the stamped capacity is ratio the two areas to the required relieving flow. People used to do this all the time with the API orifice but they were wrong in many an instances. The ratio only works if the API orifce/ASME orifice = 0.9 and it just isn't always so.

I think you mean ASME orifice/API orifice=0.9

As mentioned by Bruce in an earlier post, ASME area is roughly 10% larger than API effective area.