Pumping of gelled fluid in pipeline applications
Master thesis
Permanent lenke
http://hdl.handle.net/11250/181967Utgivelsesdato
2011Metadata
Vis full innførselSamlinger
- Studentoppgaver (TN-ISØP) [1410]
Sammendrag
This master`s thesis is called “Pumping of Gelled fluid in Pipeline Applications” and is carried out on
request by Halliburton Pipeline & Process Services. The reason this thesis is carried out is that more
information is needed to predict accurately the required pumping pressure to encounter when pumping
gelled fluids.
A test program was set up to provide data from situations similar to those from real projects. The test
program was designed so it would be easy to collect and compare data. It was therefore carried out by
pumping different gel plugs of lengths of 50, 100 and 150 meters into three different pipelines of 2, 4
and 6.1 inches in diameter.
The gel used in these gel tests is called Temblok-50™.This is a Halliburton produced gel which is made by
mixing a linear gel with a cross-linker. Cross-links are bonds that link one polymer chain to another. The
Temblok-50™ is a viscous water based gel with an extremely tough cohesive structure developed for use
where a long life material is required. It is formed by cross-linking a natural gum or its derivatives in
alkaline pH conditions. Temblok-50™ is also a thixotropic gel because it has a time dependent viscosity.
Fanning`s equation is used for calculating the theoretical pressure needed for moving gel plugs. A
rewriting of Fanning`s equation gives the equation which is used in this thesis. This equation is shown
below with an explanation of the different characters:
Where:
= pressure for moving the gel plug [bar]
= pipe diameter [in]
= wall shear stress [bar]
= length of the gel plug [m]
A thorough explanation of this equation and its use is seen in chapter 6.1, Theoretical pressure.
A test program for gathering accurate pumping and pressure data was planned. The aim with these tests
was to figure out what pressure is needed to move gel plugs in a pipeline. The different gel plugs were
placed inside 2 and 4 inch pipelines at a test site in Risavika Harbor and a 6.1 inch pipeline at IRIS
research center`s test site. To figure out how gel plugs move in the pipe, and if it is possible to get water
pushing the gel, to flow past the gel plug without moving it, transparent pipes were used in different
tests at the University of Stavanger.
Before the tests in Risavika Harbor and at IRIS were carried out, safety margins between pressure safety
valves (PSV) and calculated pressure were calculated. It was also calculated how much gel was needed.
Initial to the gel tests was a tests set-up containing the P&ID (piping and instrumentation diagram),
procedure and a work schedule designed. The test set up in Risavika Harbor was made of two main
units, one pumping unit and one pipeline unit. The pumping unit is the same for every test, but the
pipeline unit is changed between the 2 and 4 inch pipeline unit depending on which pipeline the test
3
was carried out in. These pipeline units were made by mounting together pup joints of different lengths
to achieve two loops of the total lengths of 350 meters for the 2 inch pipeline and 250 meters for the 4
inch pipeline. At IRIS`s test site there is a 700 meter long and 6.1 inches in diameter permanent pipeline
to which the pumping unit was connected. This pipeline is build by mounting together casings.
At the University of Stavanger a transparent pipe with the inner diameter of 1.57 inches was placed
vertical and a rubber plug connected to a pole was placed inside the pipe. Linear gel and cross-linker was
poured into the pipe. Gel plugs ranging in length from 30 cm to 120 cm were set to settle in the pipe for
more than 12 hours. After gel plugs had been settling in the vertical pipe for more than 12 hours the
whole pipe was moved and placed horizontally in the test set up. The rubber plug was then removed by
pulling it out while rotating the pole.
The test set up consisted of a small Gilson pump, a hose which connected the Gilson pump and the 1.57
inch transparent pipe. The transparent pipe had a t-junction connection in the beginning of it. A pipe
was placed vertically from this t-junction and it had a valve connected to the top of it. A hose going
upwards was connected to the end of the transparent pipe at the end of the pipe which was in front of
the gel plug. The reason for using the vertical pipe from the t-junction and the upwards going hose was
to make it possible to fill the pipe with water both in front and behind the gel plug, which was the last to
be carried out before a test was run.
The results produced by the tests in Risavika Harbor, IRIS research center and at the University of
Stavanger were interpreted and presented in this thesis. There are many variables which have to be
considered when interpreting the data from the gel tests. These variables are such as pipe diameter,
length of gel plug, temperature, shear stress, cracks, settling time, first and second time the gel plug is
set, flow of water pushing the gel plug, remains of gel in the pipeline and uncertainty of the data
collected.
From these variables it is discovered that all of them had some influence on the pressure needed to
push gel plugs. The variables of first and second time gel plugs are set and remains of gel in the pipeline
could be neglected when the pressure needed to start moving gel plugs are calculated. If it is the first or
the second time gel plugs are set has almost no influence on the pressure needed to move gel plugs.
Remains of gel in the pipeline have a bigger influence on gel plugs when they are already set in motion.
The variables which need to be included, when pressure needed to move gel plugs is calculated, are pipe
diameter, length of gel plug, temperature, shear stress, settling time and flow of water pushing the gel
plug. It can be concluded that Fanning`s theoretical equation does not contain of enough variables to
calculate the correct pressure needed to move gel plugs. This equation contains the variables; pipe
diameters, length of gel plug and shear stress. Experience from the test results is that this equation lacks
the variables of settling time, temperature and flow rate of water pushing the gel plug.
In the theoretical equation for calculating pressure needed to move the gel plug the influence of length
of gel plug and pipe diameter is probably right. The shear stress which is seen in the equation as has
the formula of
. This means that the formula for shear stress contains viscosity and velocity of
4
the gel plug, and pipe diameter, but none of the other variables. The influence of these three variables
in a new expression for shear stress which will be a part of a new equation for calculating the pressure
needed to start moving gel plugs in pipelines is unknown. What is known is that in this new equation
temperature, flow rate of water pushing the gel plug and settling time has to be included.
Settling time influences shear stress because the gel plug “sticks” more to the pipe wall when it settles.
This will make it harder to move and gives in a way the wall more shear stress because the gel plugs
“work” together with the shear stress of the pipe wall and provides more shear stress which the pump
has to overcome to push gel plugs forward.
The influence of temperature on the pressure needed to move gel plugs was not measured accurately
enough. What can be concluded from the gel tests is that when a pipe with a gel plug inside is exposed
to temperatures below zero degrees Celsius, the pressure needed to move gel plugs will be much lower
than if the pipe has not been exposed. It can also be concluded that gel plugs will not be destroyed by
being moved in or influenced by the pipe being exposed to temperatures below zero degrees Celsius.
Flow rate of water given by the pump does not directly influence the shear stress, but it influences the
shear stress indirectly. Too low flow rate will make the water, which is supposed to push the gel plug,
flow into the crack between the top of the gel plug and the pipe wall. This crack is made because gel has
a higher density than water and therefore water tends to flow on top of the gel. If the flow rate given by
the pump is too low, then water will flow past the gel plug on top of it. Water can then either flow all
the way across on the top if the gel or it can “drill” itself downwards in the gel plugs if there are air
bubbles or bends in the pipe. When the flow rate of water given by the pump is high enough, it will
“punch” the end of the gel plug so hard that it will be pushed upwards and seal the crack in top of the
gel plug, making it possible to neglect the influence of this crack.
The tests carried out at the University of Stavanger in the transparent pipes showed that there are many
coincidences involved in the behavior of gel plugs. It was discovered that the top of the gel plug is the
first part of it to be pushed forwards by pressure coming from the water which is pushed by the pump.
The whole gel plug therefore starts to move first in the top of the gel plug and last in the bottom of it.
The shape of the end of the gel plugs and if there are bubbles in it influence the pressure needed to
move the gel plug. If the end of the gel plug is inclined the flow rate needed to push the end of the gel
plug upwards has to be bigger than if the end of the gel plug is vertical. Because the shape of end of the
gel plug is not easy to predict, it is hard to figure out if the water is going to flow on top of the gel or
push it forwards. If there are air bubbles on top of the gel plug near the end of it the air bubbles will
influence the pressure needed to move gel plugs because the air bubbles will help water to flow on top
of the gel plug. Both air bubbles and the inclination of the end of the gel plug can cause the gel plug to
split into two gel plugs.
The conclusion is that the gel tests have given much information about how gel behaves in pipes and
different pressures needed to move different lengths of gel plug in pipes of different sizes. A complete
equation for calculating the pressure needed to move gel plugs cannot be calculated from these results
because the behavior of the end of the gel plug and the influence of temperature is not well enough
5
documented. There are still many things to be tested before an equation which calculates the pressure
needed for moving gel plugs can be constructed, but these gel tests are a step in the right direction.
Beskrivelse
Master's thesis in Industrial economics