Flow Assurance Technical Section

I don't think the apparent bulk viscosity changes. Maybe preferential wetting with oil for HDPE (vs water for steel) and resulting liquid films with different properties at the wall? Does the difference go away below 40% watercut, and is it velocity dependent?

Thanks for the feedback Richard.

I also think the wettability with oil for HDPE might be different compared to oil with steel, which might or might not (significantly) effect the apparent viscosity, but couldn't find any information on this.

We don't have data below 40% because the amount of production flowing through this pipe section in the field (crossing a river) doesn't have less water. In other parts in the field we saw a lower apparent viscosity at higher velocities (but the water cut wasn't exact the same), unfortunately we don't have the data to compare HDPE with steel pipe at different velocities values for the production flow of the same area (in practice this would mean laying pipes with different ID sizes, HDPE and steel, in the field and sending the area production through it). The difference in velocity between HDPE and steel is because the ID of the newly laid HDPE pipe (11.154") is different compared to the previous steel pipe (9.75"). 

Henk

Henk, some (idealized) data from the web:

Contact angle water-polyethylene: 96° https://www.accudynetest.com/polytable_03.html?sortby=contact_angle

Contact angle water-steel: 40-70° https://www.ohio.edu/engineering/corrosion/research/upload/8215-11.pdf

Contact angle oil-steel: 150° https://www.ohio.edu/engineering/corrosion/research/upload/8215-11.pdf

Contact angle alkanes-polyethylene: 0° (complete wetting): Persson, Bo "Sliding friction - physical principles and applications" pp.135 [can be found in google books]

So preferential wetting looks probable. Quanitfying the effect will be difficult though - you might have surface active components, your solid surfaces might not be clean etc.

Regards,

Richard

I guess you should check the friction coefficients of the different materials.

Steel has usually the lowest ones.

Then it really depends on the internal surface quality and how smooth it is.

Any other variable under control (valves, flow rate, properties of the fluid being transported, etc..) ?

Regards

Hmm! Tricky.

Presence of oil-wet particulates in an oil-water stream should shift the transition point (to water-continuous) to higher water cut; presence of water-wet ones should do the reverse.
Getting over the viscosity hump and into water-continuous generally leads to a large drop in viscosity of the stream.

So, I'm wondering whether the presence of a strongly oil-wet surface might be shifting the transition point  to higher water cut, which is showing up as a higher viscosity fo the mixture. I just cannot get my head around why that would be such a large effect with a tubing surface, which should have much lower surface area than a large number of entrained particulates.

Something to consider because it would produce an effect in the direction you are observing; I'm just not convinced it's of sufficient magnitude to account for what you are seeing.

On the other hand, you present no statistics to indicate your confidence in the magnitude of this effect. I can imagine it might be quite difficult, comparing different pipelines with different ages (corrosion effects for the steel, etc.), apparent liquid velocities, water cuts, scaling tendencies, sanding tendencies (?), etc. Not easy to de-convolute statistical signifiance in such a system without a large number of pipelines.

Thanks Dario,

Basically we have to do some fundamental lab testing to see how much impact the wettability oil steel vs oil HDPE has on inversion point water continuous to oil continuous and as a result on the apparent viscosity of the field conditions. Hoped that this was done already, since the use of HDPE in the oil industry has increased. 

Hmm! Tricky.

Presence of oil-wet particulates in an oil-water stream should shift the transition point (to water-continuous) to higher water cut; presence of water-wet ones should do the reverse.
Getting over the viscosity hump and into water-continuous generally leads to a large drop in viscosity of the stream.

So, I'm wondering whether the presence of a strongly oil-wet surface might be shifting the transition point  to higher water cut, which is showing up as a higher viscosity fo the mixture. I just cannot get my head around why that would be such a large effect with a tubing surface, which should have much lower surface area than a large number of entrained particulates.

Something to consider because it would produce an effect in the direction you are observing; I'm just not convinced it's of sufficient magnitude to account for what you are seeing.

On the other hand, you present no statistics to indicate your confidence in the magnitude of this effect. I can imagine it might be quite difficult, comparing different pipelines with different ages (corrosion effects for the steel, etc.), apparent liquid velocities, water cuts, scaling tendencies, sanding tendencies (?), etc. Not easy to de-convolute statistical signifiance in such a system without a large number of pipelines.

More suggestions:

- Larger diameter HDPE pipe facilitating demulsification at lower velocity for this particular oil and temperature, resulting in stratified flow and causing higher water holdup in river crossing low spot and higher observed dP. Restriction is not solid (as wax) but hydraulic (water) leaving less effective area for flow. Pipe oversized for the flowrate.

Less likely possibilities also considered:
- different thermal expansion / contraction diameter reduction of nonmetallic pipe

- diameter reduction of nonmetallic pipe under river water hydrostatic pressure

- possibly gelling of oil (albeit unlikely at such high water cut and at 85-90°F)

This would make an interesting case study and presentation for SPE FTS if someone were to compare a range of software tools vs field observed information.

More suggestions:

- Larger diameter HDPE pipe facilitating demulsification at lower velocity for this particular oil and temperature, resulting in stratified flow and causing higher water holdup in river crossing low spot and higher observed dP. Restriction is not solid (as wax) but hydraulic (water) leaving less effective area for flow. Pipe oversized for the flowrate.

Less likely possibilities also considered:
- different thermal expansion / contraction diameter reduction of nonmetallic pipe

- diameter reduction of nonmetallic pipe under river water hydrostatic pressure

- possibly gelling of oil (albeit unlikely at such high water cut and at 85-90°F)

This would make an interesting case study and presentation for SPE FTS if someone were to compare a range of software tools vs field observed information.