Propylene Glycol Correction Factor Question

[kaconnol]

I’m sizing a pump for a heat pump cooling water loop and have a question about propylene glycol correction factors.

I determined the required flow rate for the heat load based on properties of 40% PG (Spht and SG).

I determined the required pump head based properties of 40% PG (Friction factor based on Reynolds number which is based on 40% PG viscosity).

When viewing pump curves to make a selection, do I need any additional correction factors or am I good since I used 40% PG properties to determine the pump flow rate and head?

My understanding is the correction factors are used when your calculation uses water as the fluid. Working with propylene glycol is new to me and I want to make sure I’ve accounted for all required corrections.


Kevin Connolly, PE

http://www.TheEngineeringSurvivalGuide.com

 

[Wilbur55]

See the discussion starting on page 4 here:

http://fluidh.com/cmsAdmin/uploads/cih-taco5aselectingthepump-website.pdf

The corrections become larger as the fluid temperature decreases.

For the purposes of scheduling the pump and reviewing the submittal, stating requirements in terms of performance when circulating pure water is easier.

Communicating expectations to the balancing contractor, and reviewing the TAB report may be less straightforward.

Balancing contractors may be unaccustomed to using or documenting correction factors when measuring flow rates through balancing valves that were calibrated for water, or for pressure gauges when measuring pump heads and reporting the values in feet (of what - pure water or glycol solution, at _?_°F?).

Where you "determined the required pump head," is this expressed in terms of "feet of 40% PG solution at the density for the flowing temperature?"

This may be interesting:

http://www.taco-hvac.com/wizard_pumps.html

It is not clear to me whether the "Design head" input field at the link above is in terms of "feet of water" or "feet of fluid." The selection program might expect the user to make corrections externally if the "fluid" is a glycol solution.

 

[Wilbur55]

A preliminary look at results from the selection tool linked above suggests it may produce exactly the same head-gpm curve regardless of the fluid viscosity. It does correct the BHP curve based on fluid density. The efficiency and NPSH curves also seem to be unaffected by fluid viscosity.

Though no information on correcting the NPSHR curve was found in the link below, one might expect variation in the head and efficiency curves, in addition to BHP curve, based on the procedure here:

http://www.mcnallyinstitute.com/14-html/14-04.htm

 

[kaconnol]

My "required pump head" is based on feet of 40% PG at the density for the flowing temperature. I did a control and calculated the required flow rate and pump head for pure water at the same temperature and it was less than the 40% PG and the values looked reasonable.

Based on the first linked article I need 3 corrections.

1. Flow rate - Check
2. Piping Friction Head - Check
3. Pump Performance - Need to use the tables in the linked article.

Items 1 and 2 determine my design point, then I need to inflate those in accordance with the tables (depending on temp and %) when reviewing pump curves based on pure H2O.

Kevin Connolly, PE

http://www.TheEngineeringSurvivalGuide.com

 

[georgeverghese]

40% PPEG would be some what more viscous than water. The generic pump curves you see usually are for water and for liquids whose viscosity is not that removed from that of water. There would most likely be correction factors for the head produced, based on reference head produced with water. Ask the pump vendor for these. Once you've selected the pump, check for how this pump behaves when oyu have some temperature excursions.

 

[CarolinaPE]

Another option is to use some selection software from someone like Taco. It allows you to input the percentage of glycol, etc when running the selection.

 

[EnergyProfessional]

Ultimately you need to calculate head in ftH2O or PSI. It seems you use head of the actual fluid (PG), which isn't used by pump manufacturers. Once you established standard ftH2O or PSI, go to Grundfos website or any other pump manufacturer and they will have an online tool where you can enter the type of fluid and they will automatically correct it for you.

how exactly did you calculate friction with the glycol? You also need ot adjust for temperature. hot Glycol will be similar to standard water, but cold Glycol will have much more friction (and less efficient pump)

 

[Wilbur55]

Don't place undue confidence in what the pump manufacture's rep, or the results produced by software or web pages available from the manufacturer. The rep may have never before corrected a performance for a glycol application. The software or web pages may have been written by recent computer science grads who have no familiarity with pumping. The quality control on what was produced may yet to be done. As noted in a previous post, the pretty graph from a web page from a manufacture seemed to account only for the variation in density. The head-gpm curves for water and 40% PG seemed to be identical. The programmer seemed to omit any correction for viscosity.

 

[TD2K]

Centrifugal pump curves are in feet of liquid. A curve that says produces 75 feet of head at a certain flow produces that head be it water, 40% PG or mercury. if you've done your calculations for 40% PG, you can use that head directly to select pump.

Note that NPSHR on a pump curve is feet of liquid, not feet of water (yes, the NPSHR is determined using water but the NPSHR is feet of your liquid, not feet of water). Take a look at the vapor pressure of the solution entering the pump and the suction pressure. The difference, converted to feet of 40% PG, is your NPSHA.

http://www.lyondellbasell.com/techlit/techlit/2514.pdf

The above link would suggest that the viscosity of 40% PG isn't going to be greater than say 10 cP. I wouldn't expect any significant corrections to be needed to a pump curve.

 

[georgeverghese]

Perry Chem Engg handbook says the viscosity corr factor for head is 1.0 for liquid viscosity < 50cP.

 

[BronYrAur]

I know I'm taking this discussion off on a tangent, but I recently posted a glycol question on he heat transfer forum. See Link below. I selected a coil using water, and then I held all parameters the same except I switched to glycol. The performance of the coil was significantly less than I expected based on "correction factors" that I had always accepted. So I don't always trust those factors without understanding where they came from.

http://www.eng-tips.com/viewthread.cfm?qid=390423