Yea, this is definitely one of those things that seems like a two sentence topic ... until you try it yourself and realize there just keeps being one more step, then just one more, then ... and it keeps going lol.
Hi Brian. You are a godsend for me right now and I hope you are doing well. I am currently working on parallel pumping stations, and I am trying to plot the system curve for a station of 8 parallel pumps. I would like some guidance on how to tackle this problem so I can create an auto operation logic. Should I create a curve for each pump separately or combine the pipe frictions etc. for the whole station? small details like this would be really helpful. Thanks for the video in anycase!
My thoughts would be 1 total combined system curve if the 8 pumps have nearly identical connections, such that bringing pumps on or offline doesn't really introduce any new losses. An exception might be where you have 4 pumps in parallel in one cluster, and then you have a second auxilliary cluster some distance away - still arranged in parallel - but where adding any of these extra pumps introduces additional friction or minor losses that are a lot different than the first four. In this case, you might need to create several different system curves, one for each combination of clusters being in use (probably 3 curves in this case, if only cluster A active, only cluster B, or both A and B). The larger your plant, and smaller the differences between the runs to each pump, more likely these differences will be negligible.
@@BrianBernardEngineering Perfect explanation. You even guessed right; it is a cluster of 4 pumps. I will definitely work on it now with this information in mind. Thank you so much and I truly hope you the best of times.
Actually, we were given an equation to estimate f from Nevers' fluid mechanics book: f = 0.001375[1+(20000*e/D+10^6/Re)^1/3] It's easy to use on Excel. But I wonder if it's preferrable over Moody diagram during exams
Main thing with that equation is that it has limits to only be used within a specified range of Reynold's number, and specified range of roughness - but if you are in that range, go for it :). For iterative solutions, equation will save a lot of time compared to the diagram. But for students just starting out, getting comfortable using the diagram isn't a bad thing, even if there are other ways that may be faster, so that's why I use the diagram in this video, extra practice.
Hey Brian, what happens if I put two DIFFERENT pumps in series? For example, if pump A is rated for a head of 35 and max GPM of 8, and pump B is rated for a head of 20 with max gpm of 17. Would my entire system now have a max head of 55 with max gpm of 17? Thanks!
Yes, your understanding is correct. In series, you can still add the 2 pump curves vertically, even if they don't start and end at the same place. At 0 flow, you add the 2 pressures to get max head. But at high flowrates, your smaller pump isn't even really doing anything, just sort of free spinning getting pushed along by the bigger pump. Between 8 and 17gpm, essentially the larger pump is doing all the work and the smaller one not really helping at all. Max 17gpm also looks correct. This is a good idea for a follow-up video actually, combining different sized pumps. I probably won't get to it for awhile, but I added it to my list.
ithe characteristic curve of a system of 3 pumps in series is given by Hp=90-Q^2 , determine the characteristic curve of single pump, is the answer to this question : Hp=30-Q^2/3?
Yes, that looks correct to me. If typing I might write (Q^2)/3 with parenthesis to ensure the reader couldn't misread the exponent as 2/3. But this is just a stylistic change, not a substantive one. your version is correct if order of operations is followed correctly.
Question: if i have 2 identical pumps in series and the series pump curve is out of reach of the system curve unlike its original pump curve does that mean it will operate in as a single pump if it's curve has an operating point? also is it normal for some pump curves to not have a free delivery point at the end ?
free delivery point - this lets pump spin at maximum speed without any change in pressure - in practice this can often lead to cavitation issues that can damage the pumps internal moving parts, so it seems like following the curve all the way down to the x axis might not always be mechanically advisable. If a system curve intersects a single standalone pump, it should always intersect 2 of the same pump in series. It will intersect at a different point, but since the single pump curve is inside the series pump curve, it's impossible for system curve to hit single, and not hit series.
A good explanation for a topic that tends to be more complex than people think
Yea, this is definitely one of those things that seems like a two sentence topic ... until you try it yourself and realize there just keeps being one more step, then just one more, then ... and it keeps going lol.
@@BrianBernardEngineering yes! Then add multiple pumps in parallel, but different sized pumps. Then it becomes fun 😆 add then controls to it!
Easy to follow and kept me interested all the way through.!
Thank you!
Glad you enjoyed it!
Thank you sir!
You're welcome. Hope you end the semester on a strong note :)
Hi Brian. You are a godsend for me right now and I hope you are doing well. I am currently working on parallel pumping stations, and I am trying to plot the system curve for a station of 8 parallel pumps. I would like some guidance on how to tackle this problem so I can create an auto operation logic. Should I create a curve for each pump separately or combine the pipe frictions etc. for the whole station? small details like this would be really helpful. Thanks for the video in anycase!
My thoughts would be 1 total combined system curve if the 8 pumps have nearly identical connections, such that bringing pumps on or offline doesn't really introduce any new losses. An exception might be where you have 4 pumps in parallel in one cluster, and then you have a second auxilliary cluster some distance away - still arranged in parallel - but where adding any of these extra pumps introduces additional friction or minor losses that are a lot different than the first four. In this case, you might need to create several different system curves, one for each combination of clusters being in use (probably 3 curves in this case, if only cluster A active, only cluster B, or both A and B). The larger your plant, and smaller the differences between the runs to each pump, more likely these differences will be negligible.
@@BrianBernardEngineering Perfect explanation. You even guessed right; it is a cluster of 4 pumps. I will definitely work on it now with this information in mind. Thank you so much and I truly hope you the best of times.
Actually, we were given an equation to estimate f from Nevers' fluid mechanics book:
f = 0.001375[1+(20000*e/D+10^6/Re)^1/3]
It's easy to use on Excel. But I wonder if it's preferrable over Moody diagram during exams
Main thing with that equation is that it has limits to only be used within a specified range of Reynold's number, and specified range of roughness - but if you are in that range, go for it :). For iterative solutions, equation will save a lot of time compared to the diagram. But for students just starting out, getting comfortable using the diagram isn't a bad thing, even if there are other ways that may be faster, so that's why I use the diagram in this video, extra practice.
Hey Brian, what happens if I put two DIFFERENT pumps in series? For example, if pump A is rated for a head of 35 and max GPM of 8, and pump B is rated for a head of 20 with max gpm of 17. Would my entire system now have a max head of 55 with max gpm of 17? Thanks!
Yes, your understanding is correct. In series, you can still add the 2 pump curves vertically, even if they don't start and end at the same place. At 0 flow, you add the 2 pressures to get max head. But at high flowrates, your smaller pump isn't even really doing anything, just sort of free spinning getting pushed along by the bigger pump. Between 8 and 17gpm, essentially the larger pump is doing all the work and the smaller one not really helping at all. Max 17gpm also looks correct. This is a good idea for a follow-up video actually, combining different sized pumps. I probably won't get to it for awhile, but I added it to my list.
ithe characteristic curve of a system of 3 pumps in series is given by Hp=90-Q^2 , determine the characteristic curve of single pump, is the answer to this question : Hp=30-Q^2/3?
Yes, that looks correct to me. If typing I might write (Q^2)/3 with parenthesis to ensure the reader couldn't misread the exponent as 2/3. But this is just a stylistic change, not a substantive one. your version is correct if order of operations is followed correctly.
Question: if i have 2 identical pumps in series and the series pump curve is out of reach of the system curve unlike its original pump curve does that mean it will operate in as a single pump if it's curve has an operating point? also is it normal for some pump curves to not have a free delivery point at the end ?
free delivery point - this lets pump spin at maximum speed without any change in pressure - in practice this can often lead to cavitation issues that can damage the pumps internal moving parts, so it seems like following the curve all the way down to the x axis might not always be mechanically advisable.
If a system curve intersects a single standalone pump, it should always intersect 2 of the same pump in series. It will intersect at a different point, but since the single pump curve is inside the series pump curve, it's impossible for system curve to hit single, and not hit series.
@@BrianBernardEngineeringthank you professor,much appreciated