Hi All,

I am now doing the Analyses for the Two phase flow subjected to Slug flow. If I check as per EI GUIDELINE the Density x v2 exceeds the limit of 5000 .

1. I am checking the LOF and based on the requirement I will maintain the LOF of the system < 0.35 .

2. I am checking the Modal natural frequency and I have maintained it to 4 HZ.

3. I have done the Slug static Analysis in Caesar.

In addition to this whether CFD analysis needs to be done as in EI guideline shown it needs specialist advice for slug lines.

Kindly clarify me.

Best Regards
K.RAJESH

Verify that the ρv² used in the calculation is the actual maximum ρv², and not just the case for slug flow.

Be aware that flow induced vibrations are not really a "slug flow" phenomenon - just where you attempt to pass an unusually high quantity of fluid through the pipe.

If possible, obtain slug frequency. If it's within 20% of your pipe natural frequency, you need to fix that. If it's less than 20% of your pipe natural frequency, then a static analysis is fine. If the slug frequency is greater than the pipe natural frequency, you should consider a dynamic analysis instead (or otherwise stiffen up your pipe).

Rajesh,
My interpretation of EI guideline is that excitation due to slug flow is not covered by EI guideline. For slug flow p-roper analysis is required.
For vibration provide adequate support to maintain higher natural frequency. You structural support should also be rigid enough to dampen vibration and withstand slug force.
As suggested by Michael, if possible obtain slug frequency and perform dynamic analysis if required.

Hi Michael,
Thank you very much for your suggestion . Kindly clarify me about slug frequency which you are defining. How to get it?

We received from process that the occurrence of Slug will be once in two month.

Best Regards
K.RAJESH

Rajesh,

Process either means that you'll see one slug every two months or that you'll see slug flow, at some frequency, every two months.

Process should advise what they mean.

Hi there,

My opinion is that it is extremely hard to assess/estimate an Occurrence Frequency of Slugs for subsequent comparison against Piping System Natural Frequencies.

Slug Flow is an irregular or quite random phenomenon. I am not a CFD or Fluid Mechanics specialist, but, what I know from many previous similar projects/circumstances, is that the simulations required to find Slug Flow parameters are complex enough, depend on several key input physical/process data and include some uncertainty.

I believe that Slug occurrence frequency, expressed in << No. of Slugs per Time >> ratio might be useful as an AVERAGED PARAMETER to estimate the overall number of Slugs during System lifetime and to establish if Fatigue analysis may be required or not.

Otherwise, Slug Flow loading is typically regarded as an Impact action. SLUG LENGTH, together with Slug Velocity and Slug Density, are the key parameters to perform a Dynamic Time History Analysis. In this way, the dynamic response of the System with respect to Slug Load excitation spectra will be quantified realistically. Depending on the Slug Flow loading type, the dynamic stresses might be qualified as Occasional Stresses (in this case the Amplitude/peak values are superposed over Sustained stresses to be checked against B31 limits) or as Stress Ranges (in this case, the peak-to-peak stress ranges, e.g. two times of maximum amplitude values found by CII analysis, are considered to be used for subsequent Fatigue Analysis).

I know from my past experience that even Slug Length is a parameter that might not be easily obtained from Process discipline (e.g. in general, Process claims that in the absence of CFD simulation data, such parameters cannot be estimated...).

As a final conclusion, the actual manner of Slug Flow analysis depends on the Project Design Basis and Client/Owner specific requirements. If CFD simulation was not stipulated within the approved Design Basis, then Client/Owner cannot ask for a comprehensive assessment of Slug Flow loading effects. Typically, in such case Slug Velocity and Slug Density are used to assess the Slug Load using DLF = 2 and then the Occasional Slug Stresses are calculated using the equivalent static approach.

Regards,

If you have a slug force, then you must somehow have the slug velocity and mass. If you have a mass flow rate of liquid, then you can reasonably predict an average slug frequency.

Whether your slugging is random or developed is very much a function of pipe routing and relative rates of gas/liquid. It is a function of the Froude numbers of your system.

Slug Density is NEVER equal to Liquid Phase Density. It is typically between Vapor and Liquid phases densities. Similarly, Slug Velocity is generally higher than Liquid Phase theoretical velocity. Since Slug Force value depends on the Density and Square of Velocity it appears that taking into account Liquid Phase flow parameters, the Force value will be UNDER-ESTIMATED.

In fact, in L+V mixture phase case, Liquid Phase and Vapor Phase flow parameters are pure theoretical values, the ACTUAL L+V mixture phase flow parameters being different. In addition, Slug flow parameters are different than the long-term L+V mixture phase flow parameters.

Therefore, taking into account Liquid Phase flow parameters it's a tricky approach that may yield to non-conform results. A conservative approach would be to consider Vapor phase velocity and Liquid phase Density, but in this case the Slug Force might be excessively over-estimated.

And, finally, such elementary approach does not predict the SLUG LENGTH, which is a key parameter is any dynamic assessment.

As a practicality, I'd want to simplify it as such to achieve a conservative force, and apply a safety factor to the correlated slug length.

The actual slug density and therefore length would be a function of the fluids' miscibilities, entrainments, and agitation (Reynold's numbers, perhaps), which, like everything else, would have to be tabulated experimentally, or estimated via analog.

If you had a liquid and gas that were fantastically phobic of one another or had fantastically high surface tension, the slug density would be that of the liquid. However, such a concoction is admittedly unrealistic.

Hi again Michael,

Entirely agreed your latest statements.

However, I hope you also agree that the suggested approach ("...The actual slug density and therefore length would be a function of the fluids' miscibilities, entrainments, and agitation (Reynold's numbers, perhaps), which, like everything else, would have to be tabulated experimentally, or estimated via analog.") includes some uncertainties, approximations and assumed data that a pipe stress engineer (typically specialized in mechanics of materials & strength of materials and having only general/summary knowledge of fluid mechanics) is unlikely to be familiar with.

Consequently, coming back to the initial post of this topic, I believe that the estimation of the Occurrence Frequency of Slugs for subsequent check against System Frequency is not a realistic approach to be used by K.RAJESH...in the absence of a PROFESSIONAL Process Engineer assistance. Since such a nice circumstance is hardly to be met in many situations, I believe that Slug Load treatment as a typical Occasional load would be the recommended approach...taking also into account that the suggested occurrence "frequency" appears to be 1 time at each 8 weeks period. This normally means that it is a low risk of fatigue induced vibration failure.

Best regards and Happy Holidays!

Hi Daniel/Michael,

Thank you for your effective interaction. As discussed getting the frequency of Slug is hard to obtain.
Static Slug Analysis with DLF = 2 is also done instead of Dynamic Time history Analysis.

Can you tell What I have done is sufficient as I said in point no.01 to 03 for my system since it is subjected to (Slug + FIV) or Any additional check is required.



Best Regards
K.RAJESH

RAJESH,

In my opinion your approach should include the followings:

1) Assuming you have a 2-phase (L+V) case, you should perform a Vibration Induced Fatigue (VIF) screening assessment for Flow Induced Turbulence mechanism based upon Energy Institute methodology, Technical Module T2.2 (typically, an Excel or Mathcad spreadsheet ma be used). The Fluid Density and Fluid Velocity parameters are those corresponding to long-term Operating conditions (Mixture Phase L+V case) - Process Discipline should commonly provide these parameters.

2) Based upon the estimated Slug Occurrence "frequency" of 1 time per 8 weeks (e.g. 2 months) period, it may be concluded that the Slugging loading-unloading overall cycles number is lower than 1000 for a typical lifetime period of 25 years. Therefore, there is a low risk of Vibration Induced fatigue Failure for the piping system and no Fatigue assessment is required.

3) Slug Flow load may be assessed and qualified as Occasional Load and corresponding stresses may be checked as ASME B31 Occasional Code stresses. Slug Flow load may be quantified using te eqyivalent static approach (DLF = 2), based upon actual SLUG DENSITY and SLUG VELOCITY - these parameters should be provided, again, by Process Discipline. Be careful, using Liquid Phase Density and corresponding velocity obtained from the current Flow Rate value is generally incorrect, since Slug Density is typically lower than Liquid Phase Density and Slug Velocity is generally between the theoretical velocities of Liquid Phase and Vapor Phase (as calculated by Process). Process should estimate the actual SLUG DENSITY and SLUG VELOCITY.

4) Taking into account that you have a 2-phase flow case with occasional Slugging Regime, pipe support arrangement should be tightened enough to ensure a proper rigidity and sufficient high level of natural frequency - e.g. it is recommended Natural Frequency to be higher than 5...7 Hz limit (see for reference DNV RP-D101 or TOTAL Piping Stress Analysis Standard GS_EP_PVV_107).

Good luck!

1: If you're only seeing slug flow once per 2 months, you may be normally operating at only liquid or gas capacity. Again, I would verify that you calculated your LOF at maximum ρv², and not some averaged ρv² your process invented for a slug case.

Fatigue need not be limited to "high cycle" applications, but I agree that 180 cycles in 25 years would be an abnormal fatigue analysis. If you instead expect, say, 50 slugs for that "once" every 2 months scenario, you definitely would be worthy of fatigue analysis. (Typical cut-off is 5,000-7,000 cycles, but is really up to the end user.)

2: Nothing to add

3: Nothing to add