Bingham Fluids and Pressure drop 1

[ColourfulFigsnDiags]

Hello all, I've been away for a while! This is probably a simple question, but I don't know much about Bingham fluids.

I want to pump an oil which has a tendency to gel. I know the yeild stress of it, so can calcuate the minimum restart pressure, but what I want to know is if that minimum restart pressure is required once the fluid is flowing.

i.e. say it requires 3 bar to overcome the yeild stress of the fluid in the pipeline, and hydraulic calcs indicate a 5 bar pressure drop is required for the required flow rate. Will the actual pressure drop for the required flow rate be 8 bar or simply the 5 bar initially calculated?

Read the Eng-Tips Site Policies at FAQ731-376

 

[BigInch]

Its not additive.

Starting up to achieve any given flowrate will initially require a higher pressure, which will then begin to decrease and do so more and more as the fluid starts moving.

How much pressure you need at any given time is dependent on the changes in velocity each differential flow element down the entire length of the pipeline is making at any given time. Considering 100% pressure that required to sustain your target flowrate when at steady state, if you start fast, you might need 150% pressure, where as if you started slowly, you may only need 120% to get going, then you will eventually have to cut back to 100% pressure to keep the rate from going too high.

A good example of a Bingham fluid pipeline start up would be something like this,

Say you have a target flowrate of 100,000 BOPD and want to consider the problem of reaching it by sequencing 1 or 2 of two equal parallel pumps, each with a capacity of 50,000 BOPD. You may find that it requires two pumps in series (2x discharge press of one) to start up the pipeline and reach an initial flow of 50,000 BOPD, which after some time the discharge head required to sustain that flow flow will decrease and perhaps eventually you will probably start to have flows higher than 50,000. So, then you can start cutting back on the discharge pressure (VSD?) WHILE THE FLOW CONTINUES TO INCREASE! After some time you may stabilize at a flow of 50,000, but with both pumps running only at half rpm (this is where a VSD is really nice). When you arrive at that condition, its time to switch the configuration of the two series pumps to two parallel pumps at say 75% rpm and start ramping up the rpm and flow to 100% rpm when you ultimately reach the target flowrate of 100,000. THEN, after the flow gets going through all downstream sections of the pipeline, you might be able to cut back the rpm of both pumps to say 85% when you finally steady out at 100,000 BOPD.

A good pipeline simulation program is just about essential to do an accurate start-up and shut-down analysis of these types of pipelines.

BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[ColourfulFigsnDiags]

Thanks, I've got access to OLGA, what other information would I need about the fluid to do the calculation?

Read the Eng-Tips Site Policies at FAQ731-376

 

[BigInch]

You have to define the nonlinear shear stress. I don't use OLGA so I don't know how you would have to enter that.

Depends on the fluid and how you want to model it.

If you are not doing isothermal calculations, you will generally want to have an accurate method of calculating viscosity at flowing temperatures. To model the fluid as compressible, you will need the bulk modulus and maybe how that varies with temperature and pressure. Heat capacity and overall heat transfer coefficients of pipe-steel-insulation and the surrounding soil, water or air, if you want to model heat transfer. Vapor pressures are also handy to have around.

BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[ColourfulFigsnDiags]

Yeah I have all that other information, it is actually quite a detailed pipeline model we are working with, full thermal and transient effects allowed for, all the thermal property dependencies for the three phase flow are calculated by PVTsim (in this case I'm only intersted in the single stabilised oil phase), but it does not simulate non-newtonian behaviour, and I was looking for the correct terminoligy to use so I don't look like a goose when I ask the client for it (although they might not have it, they've only done a little experimental work on this stabilised crude) ... nonlinear shear stress ... excellent, sounds important enough to me!

Thanks again!

Read the Eng-Tips Site Policies at FAQ731-376

 

[BigInch]

Bingham-fluids actually have a yield stress, so what you really need is the details of that yield stress, how it varies with temperature and pressure. Contrary to newtonian fluids, Bingham fluids can actually transmit a shear stress without a velocity gradient (like a solid), but in order to make a Bingham fluid flow, the shear stress has to be larger than the yield stress. Below the yield stress the fluid will act pretty much like a solid body, but above the yield stress it acts as a liquid.


BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[ColourfulFigsnDiags]

Yep I've got the yeild stress of this fluid, which is (for arguments sake) 11Pa. Therefore for a 12" (ID) pipeline that is 10km long, I can calculate that you will need about a 15bar pressure difference to overcome the yeild stress. What I want to know, is that once I have got the fluid going, what will be the pressure drop ... eg if the fluid was newtonian I calculate that the dp will be 4 bar at 50kbbl/d can I say that the non newtonian fluid with the yeild stress will require 15+4 or 19bar to make it flow at 50kb/d? And I think in your first post you said that I cannot say that.

Read the Eng-Tips Site Policies at FAQ731-376

 

[BigInch]

What's the fluid???

You'll need not only the yield stress but a plot for stress vs velocity. Its a rate of shear stress. A Bingham fluid has a yield stress then the rate is constant, diliant fluid increasingly curved up, pseudoplastic slopes up then flattens out. Some fluids have different plots based on the time they have been flowing, or what they've been doing for the last week or two, so you must have all changes with whatever dependent variable of relavence well defined.

If you need a paper on this, I've uploaded one for you, as its getting a little complicated to continue discussing in Tips...

Get it here,

http://rapidshare.com/files/7169944/Non-Newtonian_Flow.doc.html

BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com