Benchmark Dataset No. 15

If I were teaching in a university (and yes... I'm looking), my very first lecture would include the topic of 'compensating inadequacies' in modeling. What's that, you might say? This refers to the way models are put together and a common feature that occurs (and can be shown to occur) in inadequately tested/validated models - that is - many individually wrong constitutive relationships/sub-models can be put together in a way that still 'match' the global tests the overall model is subjected to. We say that there are compensating inadequacies in the overall model in which a wrong sub-model in one part can compensate for the error or degree of wrongness of another sub-model in another part of the overall model. Many scientists or engineers who are unaware of this see this as a remarkable flexibility of models. But it is not.

What is really going on is poor testing, or inadequate testing of the model. That brings up another related topic of global testing versus specific (or special) effects testing of models. If there are carefully designed experiments that measure (or isolate) specific effects of a model, then one can show how each constituent part of an overall model can be individually validated. This way, it is easy to show how that individual variable or constituent relationship governing the variable can directly lead to bad predictions in the overall model. We see this all the time in multiphase pipeline liquid inventories, as an example. It is not uncommon to find many different multiphase pipeline simulators predicting similar correct pressure drops (within acceptable accuracy ranges) but are very different from each other when it comes to liquid holdup predictions along the pipeline (e.g., go ahead and model for yourself the multiphase subsea pipeline Dataset No. 9 with your own simulator). The global macroscopic, integrative variable in that case is pressure (it includes the integrated effect of everything) and the dependent variable of liquid holdup along the pipeline is the specific local variable that contributes to the pressure calculation.

So why all this prelude? I've highlighted above to show that when you are evaluating the capability, accuracy or adequacy of the multiphase flow simulator you are using in your projects or daily work, you also have to look at your local predictions in addition to your global predictions. You have to be able to gain confidence in each constituent part or sub-model of your overall model. In the benchmark dataset I highlight today, we do this by looking at the gas-phase shear stress in stratified gas-liquid flow (an often encoutered flow pattern in gas condensate pipelines), which is a local variable that contrbutes to the bigger calculation of wall (or total) shear stress that is needed for your erosion/corrosion/friction calculations. The latter is itself a sub-model that contributes to your frictional pressure gradient calculation. So we are testing a sub-model of a sub-model. That is really good testing. It allows us to independently evaluate with little doubt that the multiphase flow model being used actually captures the fundamental understanding of the local phenomenon in addition to how that local phenomenon interrelates with other local phenomena and how they all contribute to the global integrative calculation of the macroscopic variable we call pressure.

Thus, continuing on the multiphase flow benchmark dataset series, here is Dataset No. 15. The modeling goal for this field dataset is to generate the flowing gas-phase shear stress and liquid holdup of a very slightly up-inclined tube (+0.104 degrees from horizontal) operating in stratified flow. Measured (not back-calculated) gas-phase shear stress and liquid holdup data are available in the original reference for comparisons to predictions.

For any multiphase flow modeler new to this benchmark series, refer to the original article here for context and general information.

Enjoy!

#phaseshearstress #erosion #corrosion #wellboreintegrity #pipelineintegrity #wallshearstress #flowassurance #compensatinginadequacies #globaltesting #specialeffectstesting #multiphaseflow #digitaltwin #benchmarkdata #digitaltransformation #digitaloilfield #analyticalmodeling #bigdata #nagoobenchmarkseries #wellbores #pipelines #flowlines #subsea #offshore #onshore #unconventional #conventional #FBHPsimulation #liquidsloading #annulusproduction #annulusgaslift #PIPsimulation #wellfluidlevelsimulation #downholeseparation #flowlinepressuresimulation #flowlinevolumesmodeling #wellboreslugging #flowlineslugging #heavyoil #lightoil #volatileoil #gascondensate #wetgas #artificiallift #productionoptimization #productionengineering #operations #facilities #virtualmultiphaseflowmetering