PROCESS ENGINEERING POP QUIZZES
We set up Process Engineering Pop Quizzes on LinkedIn in various forums, including Process Engineering, Oil & Gas Professionals, AICHE, GPA Rocky Mountain, and others. Follow our LinkedIn button at the bottom of the page and check out the past ones. We will compile all of the quizzes over time and publish an e-book with these quizzes for use in academia or in training for workplaces.
Feel free to comment on the LinkedIn forums and help others out. We post the answers after a few days of commentary allowed by everyone.
We will set up the pop quizzes here on this website over the next couple of months and redirect conversations here. Thanks for the participation.
NEWEST POP QUIZ 18-4 HERE: Process Pop Quiz Answer
Quizzes for September have not yet posted
Quizzes for October have not yet posted
Quiz 11-1--Dynamic Flow in Low Pressure Gas Line Shown Below Here:
Month 11, Quiz #1.
So I have a real world problem this month and I truly have only a guess on how to solve it instead of really knowing how to solve it. Without providing details, this is a very important issue to solve besides just academically.
I have a 8" sch. 40 line running about 1200 ft. through a pipe rack and injecting gas into a flare header (call this tee "Node 2"). If the line is not flowing at all, and then I open a valve and BEGIN to inject gas at 0.05 MMSCFD (million standard cubic feet per day) into this 8" line 1200 ft. away at Node 1, how long before the system starts ejecting gas into the flare header on the other end? Will it be sudden, or slow ramp up? Linear or s-curve flow vs. time?
Operating Pressure of the line is 14.7 psia before it starts flowing gas into the line. At this size line and this small of flow rate, the pressure drop at STEADY STATE conditions is only 0.0016 psi drop (negligible, but it's a data point).
You may assume at time = 0 seconds, that the line is already inventoried with the same gas (methane) that is being injected at Node 1.
So I'm not looking for residence time of the gas in the piping (that's trivial)--but the time it takes for injection to start at Node 1 before you start seeing steady state flow at Node 2 at low pressure operation.
Same question--what if the flow is 2 MMSCFD?
I've done some calculations with a friend with some simplifying assumptions and got a really small number for the 0.05 MMSCFD rate. Considered doing a dynamic simulation using VMG, but haven't gone there yet. Let me know how you would approach this topic.
ANSWER: So I knew the answer was fast, but was not confident in how fast. I initially did a calculation on the change in volume on the system as a “first pass guesstimate” with some help from Chris Ewasko (to correct my math) and we got about 0.1-0.2 seconds that the flow would start at Node 2. However, this is a momentum transfer phenomenon and dynamic, and by posting this topic I was seeking a better answer.
I remembered using VMGSim Dynamics in my previous work (See their Linked in profile at Virtual Materials Group) and called them up for help. VMG’s technical staff walked me through how to easily set up this problem in VMGSim Dynamics and produced the results in the pictures shown. We changed the problem up very slightly and assumed we had steam at a small rate running through the flare header in the 24” line at Node 2. For simplicity we monitored the heating value of the fluids at Node 2 mixing as verification that flow was entering the flare header at Node 2. This shows that nearly steady state conditions started within 30 seconds or starting the operation and average flow results were reaching “nearly” steady state conditions within 5-10 seconds--but notice how flow started injection to the flare header almost immediately (analogous to the 0.2 seconds determined above). What’s far more interesting is the pressure wave reflections, as you’ll see negative flow achieved for < 1 second at the early stages of flow being established, but overall, the flow is positive at Node 2 and sometimes greater than the flow entering the pipe. The gas is obviously “pulsing” the system as it starts moving.
A great thing to watch to see this actually occur is flare stacks…I’ve witnessed them “pulsing” like this at low flow conditions…you’ll see the flame “bounce” up and down or pulse up and down the initial moments it starts to flare, and the it lines out after a bit. High flow rates—not as noticeable.
I was amazed at how easy it was to set up the VMGSim Dynamics model (Nick and Gerald were doing it for me)—but I have used it before on launching pigs down pipelines models and slug sizes, and that’s why I called them to set something like this up for me. Very informative and had a TON more options and details available to model that I’m not discussing here.
The pictures are worth a 1000 words. Enjoy them and watch next week for another pop quiz.
Legend of colors: Black flat line at bottom represents flow of gas from Node 1 at steady state conditions into the pipe.
Black “wavy” line is total mass flow in the 24” header mixing with the steam flow and represents total flow (steam plus methane in the 24” header).
Green line is the net heating value of the mixture of fluid down the header (also oscillating).