Viscous Fluid (Fuel Oil/Diesel Fuel) Pipe Sizing

[PEDARRIN2]

I am trying to set up a spreadsheet to obtain friction loss for fuel oil/diesel fuel fluids - or fluids with viscosities much in excess of water.

When I use D'Arcy, I find that there is a fairly narrow band between where the flow would be laminar Re<4,000 and where the velocities are getting a bit high.

A couple questions.

1. What are the recommended flow velocities I should be shooting for. I have seen information that says 3-12 feet per second. The fluids are going to be clean, for the most part, although they can start to degrade if they stay in the storage tank too long.
2. Is D'Arcy the equation I should use. I can iterate the friction factor so that is not an issue, but since these flows extend into both the laminar and turbulent regions, the friction factor calculation would be different in both. I have seen reference to Churchill as a means to perform this calculation, but have not used it much. Would this be a good application?

Thanks in advance for any help.

 

[zdas04]

This is a really messy topic. I see a lot of crudes, many diesels (i.e., fuels with really high paraffin numbers), and all emulsions that exhibit strong non-Newtonian characteristics (i.e., weird stress/strain curves). D'Arcy (and all of the simplifications based on it) absolutely requires Newtonian (i.e., linear stress/strain curves).

As long as you are confident that the fluids are Newtonian, D'Arcy is a good choice. If you have any concern about non-Newtonian behavior then you are pretty much on your own. None of the mainstream correlations come very close to matching measured data.

David Simpson, PE
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[BigInch]

Suggest you do your spreadsheet for gasoline, diesel, water, etc. Heavy fuel oils and crudes can vary significantly (cP to 300-500 or more) and they may require some additional knowledge about how they act outside normal temperatures, (ie. how their viscosities act towards extreme temps) and how their composition affects flow in terms of wax thickening with temperature drops or asphaltine deposition. Keep that in mind and your spreadsheet should be fine. For heated crudes and heavy oils, Shell-MIT is a good optional flow equation. I always target 3m/s as a maximum long pipeline flow rate to determine initial flow, diameter, pressure drop characteristics, but that depends on if you are working with short pipes, or long pipelines. Short pipes can have velocities to 5-7m/s, whereas to limit pressure drops to affordable values on longer crude pipelines and to limit pump stations to reasonable distances apart, you usually want to look at 2.0-3m/s, using lesser velocities with higher viscosities liquids.

 

[bimr]

Is this application for a standby generator? If so, the supply piping is typically sized for 7 ft/sec as diesel may foam at higher velocities. Return piping is one size larger than supply.

 

[zdas04]

bimr,
That PowerPoint certainly was more cut and dried and confident than I've ever been. It felt like they were treating "rules of thumb" as "laws of nature". I don't mean to disparage rules of thumb, but they are just techniques that have worked in some number of cases and are generally a good place to start. That document reads like it is the only place to end. I'd be leery of guidance that was that cut and dried for all cases that ever were or ever would be.

David Simpson, PE
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[PEDARRIN2]

The application is mainly for standby generators, although it could be for HVAC and/or plumbing boilers.

I have heard that exact presentation by Mr. Drow. His presentations typically are very knowledgeable, but I always like to see the engineering behind the rules of thumb so I can know when to not use them.

A problem I see as I am developing my spreadsheet, is that if I limit the flow to less than 7 gpm, most of the flows are going to be in the laminar, not turbulent with Re <3,000-4,000. In laminar flow, the friction factor will be 64/Re which is much different than what would be calculated by the iterative approach used for D'Arcy so my numbers are going to be much different. And in the "critical area" between laminar and turbulent, I have read that all bets are off.

However, I have also read on this site, that the use of Churchill's equation mitigates these disparities. And my spreadsheet calculations seem to bear that out. In the laminar area, Churchill gives me similar results as 64/Re and in the turbulent region, it gives me similar results to D'Arcy. In the middle regime, it appears to be about in the middle of the other two approaches.

These are all numbers to me - as I have no real experiential basis for either approach. I also understand the spreadsheet will be a tool and has to be used as such, as a guideline, not necessary something written in stone.

 

[BigInch]

It's a pretty specific application (Diesel Day fuel system) and looks quite well adapted to suit it... IMO.


BTW, Churchill will also fit well.

 

[bimr]

Preferred Utilities also has a pipe sizing program. Have you tried it and if so, how does it compare to your spreadsheet results?

 

[katmar]

I use the Churchill equation all the time and I believe it really is a work of genius. But there seems to be a bit of a misunderstanding of where it fits in relation to the Darcy-Weisbach equation. Churchill is not an alternative to D-W. Churchill is a way to generate friction factors that span the full range of Reynolds numbers. This means that you use Churchill to calculate the friction factor, and then you use that friction factor in D-W to calculate the pressure drop.

Churchill uses the standard 64/Re to calculate the friction factor in the laminar regime so the results should be very similar to those you get using 64/Re directly. The main benefit of the Churchill equation is that it is able to combine separate equations for the laminar, critical and turbulent regime friction factors into one continuous equation. This makes it ideal for computer use where gaps and overlaps (as are often shown on Moody charts) would cause an automated computer solution to fail.

Churchill does make recommendations for formulas to use in the critical and turbulent regimes, but you are not constrained to using his recommendations. For example, in the turbulent regime I use the Colebrook-White equation and then use the Churchill method to combine it with the other regimes into one continuous formula. I do not believe that Colebrook-White is particularly better than Churchills equation for the turbulent regime, but C-W is the yardstick that all others are measured against and the results will match results from the Moody chart (very slightly) better than Churchill's does.

You are already aware of the problems in the critical regime. One downside of Churchill's equation is that because it generates unique friction factors in this regime people start to believe that they are accurate. The problem in the critical regime is that flows are not stable and the friction factor will vary with time. No method can predict accurate friction factors in the critical regime because they simply do not exist. The benefit of Churchill is that it generates unique friction factors in this regime - making computer solutions more stable. The best bet is to stay away from designing for flows in the critical regime. Rather change the pipe size and force the flow into either the laminar or turbulent regime.

Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[zdas04]

Katmar,
Excellent description of the friction factor discussion. What you see as a "feature" of Churchill, I see as a "bug", but that is perspective. I've always followed the advice to design as far away from the transition region as I can. Trying to assign friction factors in a rapidly changing flow regime leads to apparent answers that fail to match measured data and can even lead you to a decision that you later regret. This is similar to the results that we get from all of the multi-phase flow correlations--the answers look so pretty on a 4-color 3D graph, but fail to match field data.

The reason for wanting to know friction factors is to make design decisions or to conduct troubleshooting analysis. Both of these activities require some level of confidence in the tools being used. I have no confidence that I can predict an order-of-magnitude friction factor in the transition region.

David Simpson, PE
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[katmar]

David, I think we are in total agreement. Please re-read the last paragraph in my previous post. I think I made it clear that I agree with your policy of staying away from the critical flow regime. But it is such good advice that it was probably worth repeating anyway!

Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[zdas04]

Katmar,
We are in agreement, the only difference is that I won't even calculate a number in the transition region since I too often see people use a number because they have it even if it is complete nonsense.

David Simpson, PE
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[BigInch]

Fortunately in industrial piping there isn't much incentive to have anything other than turbulent flow, unless daddy owns a pipe supply company. It's nice for aquariums though.

 

[PEDARRIN2]

Katmar,

You are correct. When I stated about the difficulties with D'Arcy, I actually meant I was having difficulties with the Colebrook formula (iterative) for determining the friction factor to put into D'Arcy. In my spreadsheet, I am using D'Arcy to determine pressure drop for the various pipe sizes and flows, using Churchill to determine the friction factor for all regions.

But I will also take the advice and stay away from the "critical zone" with the pipe sizing.

 

[BigInch]

Well good idea, but it's not always so easy. Hot heavy oil pipelines can run in the turbulent range in some parts and laminar in others. Yes with the same diameter and at the same flow rate. Pressure drops are nonlinear varying with changes in viscosities that can vary with both temperature and in the case of non-Newtonian flow, with shear rates. In fact, in some cases you can have laminar flow near the pipe wall and turbulent flow at the core.

 

[zdas04]

I think we've come full circle in this discussion now. Disregarding non-Newtonian factors in waxy hydrocarbons is often a mistake.

David Simpson, PE
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[BigInch]

We've got over 30% wax. Very difficult to disregard, if it cools down.

 

[PEDARRIN2]

If I am dealing with diesel fuel for a generator - I am assuming I should not have to deal with non-Newtonian issues.

I know it can become a "gel", i.e. not able to be pumped at lower temperatures, but is it becoming non-Newtonian as it approaches that temperature?

If so, how should I deal with it, other than specifying some type of heater in the tank, similar to a diesel vehicle?

 

[BigInch]

Yes it will go nonNewtonian at low temps.
Heating it brings it back.
Heating tanks, heat tracing pipe, keeping it circulating and using insulation will all help.

 

[bimr]

Diesel fuel gelling occurs when the paraffin present in diesel starts to solidify when the temperature drops. At 32 degrees F, the wax in liquid form will crystallize and leave the fuel tank clouded. At 10-15 degrees F, the wax will start to gel and clog the tank and fuel filters.

An anti-gel additive will keep the diesel fuel flowing well in colder temps and will keep the diesel stabilized.