Modeling Booster Pump

Question

Hi, 
 I'm modeling a pipeline that has a booster pump in the middle (something like the scheme below) and I would like to know if there is any inconvenience in separating the pipeline and creating two models to analyze the behavior when the pumps stops. I know I will have to apply the boundary conditions as if it were a single pipeline) 
 I ask this question because we do not normally model booster pumps / stations.

Jesse Dringoli · Accepted Answer

Joao, 
 Bentley Technical Support cannot provide consulting services. It is up to you as the engineer to decide on the surge mitigation strategy, based on your judgment and knowledge of the system and standards, budget and other constraints. We are happy to assist with any specific questions you have about the use of, or assumptions behind our software. 
 With that being said, the transient simulation in the model that you sent is showing significant negative pressures and vapor pocket formations at the downstream end of the system. This is due to the uphill profile and you can read about some of the related challenges with uphill systems here . 
 You mentioned that you have three hydropneumatic tanks in the model but I only see two that are active (HT-1 and HT-2-A), in the scenario that was set active ("WHT") 
 You may want to consider adding a hydropneumatic tank on the downstream side of pump "BS". I also noticed that the transient impact is minimal on the left side of the profile between the two pumps, so you could also test removing the two other hydropneumatic tanks. Furthermore, you could consider additional scenarios to test the impact of other surge mitigation strategies such as pump flywheel (higher inertia), bypass arrangements, check valves, surge valves, etc. 
 As an example for illustrative purposes, if I set only tanks HT-2 and HT-2-A as active and multiply the inertia for the first pump by five, the transient envelope looks much better when compared to the model as you sent it, with only an air pocket at the downstream end of the system (at the D2A, which is unavoidable). Again though, this is just a quick example, and there may be many other possible surge mitigation strategies that meet your needs. 
 You should also remove the check valve on pipes PMP1-A and BS - H2 as you have already specified the check valve option in the pumps adjacent to these pipes.

Sushma Choure · Answer

Hello Joao, 
 Here is information about modeling booster pump in water products, please go through it, its similar to normal pumps. 
 How do you model a booster pump in WaterGEMS/WaterCAD?

Jesse Dringoli · Answer

Joao, 
 I agree with Tom's comment about how you should keep both lines in the same model. HAMMER can be very sensitive to modeling artifacts and approximations - it is important to "model what you've got" for best accuracy. I see that you uploaded a model, but did you have some specific question you need us to answer? Generally speaking it is expected for transient waves to interact and they can indeed combine to produce higher high pressures and lower low pressures. You can read more about that in our Advanced Water Distribution Modeling and Management book (chapter 13) and in the formal HAMMER training course listed here .

Modeling Booster Pump

Top Replies