Electronic Pump Curves

[IanVG]

I've been using the software Pipe-Flo to build and analyze piping networks. A part of their program allows you to input pump curves supplied by the manufacturer. I was wondering - is this a feature that is unique to the software I am using, or is there is a standard/convention for electronic pump curve data? A quick internet query didn't really help.

 

[HVAC-Novice]

I would say this is sunique.

But you would have to enter all the manufacturer's models that are interesting for the type of system. So your approach is only useful if you already know what pump family you want.
You are better off to use the design data (flow/pressure) and use the pump manufacturer software. That way you get all the applicable models they have and you may not know about.

Pump manufacturer software also accounts for viscosity and selects the specific motor and determines efficiency, bhp etc. I'm not sure if your piping system software does that, or how much data you would have to enter. I assume you have to enter everything manually.

Do you enter the piping system in a 3D graphics, or manually?

Are you designing for industrial processes or buildings? If buildings, I'd say most people would be better off doing all that in Revit. That way you don't enter the whole piping system and all equipment manually.

 

[MintJulep]

I guess they want to steer you to their Pump-Flo platform.

 

[IanVG]

To respond first to HVAC-Novice: Thanks! I suspected it was a semi-proprietary feature of the software I was using. And yes, to be clear, I am modeling the chilled water districts at a large campus, so all the pumps I am interested in obtaining these "digital pump curves" are existing pumps. The software I am using (Pipe-Flo) does account for viscosity and other temperature/pressure dependent fluid characteristics. And yes, for pumps that are not included in the software, I have to manually input point-by-point the head/flow points along the curve; the same manual input process goes for the BHP and NPSHa curves for the pumps. Haha, to be generous, it is a menial and time-consuming method.

I am designing chilled water districts (i.e. piping networks). I don't trust Revit to do what I am doing. I have too many problems just doing simple pressure analysis stuff in that software. It's too easy to mess-up and too difficult to quickly identify what's messed up. In addition, I wasn't aware that Revit could do pump curve stuff, I was just aware of the pressure drop calculation features.

MintJulep: Yup, that's what I suspect. Although considering, as far as I am aware, that no other piping network software can do this, I don't mind this kind of software environment shepherding .

 

[HVAC-Novice]

Just to be clear, REvit does NOT deal with pump curves. All you get is a pressure based on the critical path. That pressure and flow is then used in the pump selection software.

I assume for a campus system you just assume each building to have a certain pressuredrop and you ar enot re-analyzing each building. i assume whoever designed the system inside the building has done that. So yes, Revit in your case may not be the best tool.

 

[IanVG]

HVAC-Novice - gotcha! Yah, okay we are on the same page then . For my particular campus system, we are unfortunately missing the 'secondary' portion of the primary-secondary-tertiary system, so the building pumps are having an effect on the primary side of the system. And, yes, I assume the system inside each building was analyzed to select the building pump.

After doing some research on possible matlab/R scripts that would allow me to use printed pump curves to created fitted curves, I stumbled on a really neat tool called WebPlotDigitizer. I think it'll do just the trick to create a fitted curve for a pump curve (especially if my pump has an impeller diameter that's not explicitly drawn!). I'm thinking a 3rd order polynomial will just fit good enough, and will at least prove sufficient to replicate the pump over in the Pipe-Flo software. There's a thread somewhere on this forum that discusses the most appropriate fitted curve for pump curves and I think the conversation landed somewhere between a quadratic/second order polynomial and a fourth order polynomial. The post is thread407-151369.

Honestly, I'm surprised I haven't used this tool before; previously, I was just printing out the charts and using a ruler to get the data I needed!