Professional Documents
Culture Documents
Ewdard Owen Work
Ewdard Owen Work
In the quest of energy resources, oil exploration and production is in its peak activity during the present era. Oil and
gas fields worldwide are reaching maturity, pushing operations to more challenging areas. During transportation of
oil through a pipeline, when the oil temperature is higher than the pipe wall temperature, there will be a dissolved
wax concentration gradient between the bulk oil and pipe wall. Crystallization and deposition of waxes occurs if the
wall/interface temperature falls below the wax appearance temperature (WAT) of oil being transported. It can affect
single wells along with transportation pipelines that are critical to the safe supply of oil to processing facilities. Wax
deposition is a very complex phenomenon. As remedial costs increase with decreasing production, wax precipitation
and deposition significantly influence the economy for a field. There are several wax deposition models with
different approaches on modeling wax deposition. The basic wax deposition models are Rygg, Rydahl and
Ronningsen (RRR) model and the Matzain model. The important element is to illustrate how wax deposition models
predict wax build up. The present study intends to analyze the effect of wax deposition on pipelines using the RRR
and Matzain models and estimate the temperature profile in the entire length of pipeline using OLGA simulation
tool. The study used experimental and field data obtain from a hypothetical field in Niger Delta to compare the wax
deposition models. The trend prediction from the Matzain model gave a much higher prediction of wax thickness
than that predicted by the RRR model. The over prediction of the wax deposit thickness of the Matzain model is
attributed to incorporate deposition enhancing coefficients introduced into the Matzain correlations. The result of the
study shows that a high deposit thickness was predicted by the Matzain model with a value of 3.80957mm at
1198.8m (red line) and 0.870068mm for the RRR model at 737.601m (black line) in figure 4.4.
NOMENCLATURE
PT pressure
TM temperature,
METHODOLOGY
3.1.1 OLGA®
The OLGA® dynamic multiphase flow simulator models time-dependent behaviors, or transient
feasibility studies and field development design. Dynamic simulation is essential in deep-water
and is used extensively in both offshore and onshore developments to investigate transient
behavior in pipelines and wellbores. The OLGA® Wax module calculates the deposition and
transport of wax components along the pipeline. It models the effects of increase in pipe wall
roughness, decreases in pipeline diameter, and the increased apparent viscosity of the oil phase
with precipitated solid wax particles. The Wax deposition module supports tuning fluid
properties related to molecular diffusion, dissolution, shear related wax transport, and effective
viscosity of an oil/wax mixture to dynamically model wax deposition, dissolution, and transport
effects. The OLGA simulator also simulates pigging operations for wax layer removal and
transport. OLGA is a multiphase flow simulator that has been widely used for several decades in
the flow assurance industry to study and predict wax deposition processes in the hydrocarbon
pipelines. OLGA is structured into modules and some of these modules include the slugging and
wax deposition modules that are commercially used for wax precipitation and slugging
prediction and calculations in the oil and gas industry. OLGA software was used in this research
In order to construct an OLGA model, it was necessary to gather data (e.g. pvt file and wax file),
to build the model and define the simulation case, and to run simulations and view results in the
form of graphs. Wax deposition simulations performed in this work was done with OLGA
2017.2.0.107 version. OLGA receives the crude oil propriety input values (for example, the
enthalpies, heat capacities and thermal conductivity) in pressure and temperature values. These
properties enter the OLGA simulator as a tab file created from the tab generating a PVT package.
The wax deposition module in OLGA further requires details about the wax component,
structure, porosity, etc., converted to a wax file in a tab format generated from the multiflash wax
interface. The wax file provides information about the wax fraction as a function of the wax
forming components, temperature and pressure, and wax mixture. Results and prediction of the
OLGA simulator are largely influenced by the accuracy of table values generated from
3.1.2 Multiflash
Multiflash is a powerful and versatile system for modeling physical properties and phase
All the thermodynamic and transport properties needed for engineering studies.
Flash calculations to determine the phases present at specified conditions of pressure and
Modeling solids formation, including pure solids, halide scales, hydrates, waxes and
asphaltenes.
aspects of a study. Each configuration can be saved for future use with Multiflash or
The fluid characterization was done through multiflash, which is a tool available in OLGA.
Multiflash works as a fluid characterization environment, where different files are generated.
Such files are later read by OLGA to determine the properties of the fluids involved in the
simulation.
The oil API density, single carbon number (SCN) composition of the oil and n-alkane analysis,
presented in Tables 3.1 and 3.2, were entered in the Single Fluid characterization option of the
PVT Analysis module as basic information needed for Multiflash to perform the fluid
characterization.
Figure 3.2: Multiflash procedure flow diagram
3.3 Crude Oil Properties for Wax Table and PVT Table files
Wax table file (. wax) and PVT table file (.tab) were the two most essential table files required to
run wax deposition simulation study using OLGA. These files contained all the properties of wax
forming components and PVT parameters. In this study, a fluid package called Multiflash 6.1
version was used to generate the table and wax files. Multiflash is a tool embedded within the
OLGA software, which is designed specifically for fluid characterization using a thermodynamic
wax model known as Redlich Kwong Soave Advanced (RSKA) model. Similarly, Multiflash
was used to understand the waxing behavior of crude oil and predicts both wax appearance
temperature, the amount of wax precipitated at different temperatures and pressures. The input
parameters required to flash a sample crude include the oil compositions and its properties, e.g.
the n-paraffins pseudo-components i.e. those with 15 carbon number or higher (defined as the
main forming wax fractions in the fluid model). Other parameters are the density of oil, wax
content, API and WAT (Table 3.1 and 3.2). Also, the files required for OLGA simulation was
Properties of the pipeline material corresponds to steel with a heat capacity of 500 J/(kg-K), heat
conductivity of 43.25 W/(m-K) and density of 7850 kg/m3. These properties are included in the
software library regarding structural-material properties. In the same library, the pipeline
thickness is defined according to the information presented in table 3.3. With the pipeline created
in the software interface, pipe length and roughness were established based on the information
provided in table 3.4. Finally, the number of sections in which the pipeline is split to increase the
Before simulating wax deposition along the pipeline, the fluid was characterized with a total wax
content of 11.03wt%, using the SuperTRAPP model for both viscosity and thermal conductivity
and LGST model for viscosity. – a fluid modelling package, which gives the fluid properties,
phase behavior and single carbon number (SCN) distribution. Two feed files containing these
properties data were generated with multiflash, which are essentially required for simulation of
OLGA software and multiflash were used in this research to compare the Matzain and Rygg
Rydahl and Rønningsen wax deposition models. The experimental data for wax deposition
simulation in the flowline was obtained from a hypothetical field. The resulting data was
processed by multiflash program in order to gain data file processable by OLGA software to
Heat
Conductivity, k
Material Thickness (m) Density, ρ (kg/m3 ) capacity, Cp
(W/m-K)
(J/kg-K)
Steel 0.0118 43.25 7850 500
Enamel 0.006 0.6 1465 2115
The tab file generated from multiflash with tabulated properties of the oil given in tables 3.1 and
3.2 respectively was imported to OLGA as a PVTFILE and wax file. A 14inch subsea flowline
was modeled to transport the crude oil from a typical production wellhead represented as a mass
source called Oil Source at a flow rate of 14000 BPD (24.6559kg/s) with an inlet temperature of
70°C in the first node of PIPE-1 into the flowline as shown in Figure 3.3. The oil flows out of the
flowline at 10bars and 20°C through a pressure node called the separator where mass and energy
interaction are represented by pressure condition of the node. The flowline also consists of a
closed node with no mass and energy transfer on PIPE-1 of the flowline.
The simulation case time was set at 100 days. The time setup also requires setting maximum and
minimum integration time step to control simulation time step. The maximum and minimum
In addition to the pipeline components, an inlet source and outlet nodes were created in the
software interface. In these components, the inlet temperature, fluid mass flow and outlet
pressure were set according to the information presented in table 3.5. When wax deposition
process is analyzed in OLGA, heat transfer calculations cannot be performed based on a given
overall heat transfer coefficient (UGIVEN). Instead of this, the inner wall and ambient heat
transfer coefficients was provided to the software in order to run wax deposition simulations.
A phase envelope was developed to understand the phase changes (gas/liquid/solid) in the
system at different temperatures and pressures. The phase envelope curve as shown in figure 4.1
was obtained using multiflash; this reveals a critical point of 555.331°C and 263.566psig. Liquid
fractions are present in the hydrocarbon mixture at all temperature conditions below this critical
temperature. This suggests the presence of heavy fractions and the stable properties of the liquid
phase at high temperatures. The phase envelop diagram in figure 4.1 showed that the fluid
system is liquid phase dominant. However, below 56°C– the wax line (blue line, point D), which
correspond to WAT value of the oil sample; the system experiences solid wax deposition
problem. A narrow two-phase (liquid and gas) region was observed, which skewed to the right
between the bubble and dew point curve. The narrow area could be due to the nature and the
properties of the oil sample. On the other hand, the gas-phase is present as a dissolved gas and
The wax precipitation curve represents the amount of solids precipitated from the fluid mixture
as a function of temperature. Figure 4.2 shows a plot of wax mass percent of liquid against
temperature. It was found that wax precipitation starts at 139°F. The mass of wax precipitated
was observed to be increasing as the temperature further decrease. The figure was plotted by
multiflash software.
Figure 4.2 Wax Precipitation Curve for the crude sample at different pressure
The wax appearance temperature was 49.784ºC at 100psig. This was calculated in multiflash
using the Cross Polar Microscopy method which assumed that a minimum of 0.045% mass of
wax must be present in order to calculate the wax appearance temperature. The amounts of wax
precipitated at temperatures below the WAT is significant, and the wax phase boundary
predicted shows that at any given pressures and temperatures below the WAT the system is
within the risk area of wax deposition. The same WAT value was observed as the pressure
increased from 10 psig to 200 psig, which implies that the WAT of the crude oil sample was only
affected by temperature change. This behavior could be due to the highly waxy nature of the
crude oil, and perhaps the amount of the dissolved gas presence is relatively small.
The single carbon number (SCN) distribution provides information regarding the total amount of
hydrocarbon components in the crude oil, including both n-paraffins and non-n-paraffins. For a
typical crude oil, only n-paraffins precipitate, whereas, the non-paraffins does not due to their
higher melting point than the n-paraffin component. Figure 4.3 shows a comparison between the
calculated single carbon number (SCN) distribution in crude oil sample by multiflash and the
experimental results. The experimental results were sparsely distributed around the experimental
Figure 4.3: Single carbon number (SCN) distribution of the crude oil sample and the
experimental data
For the flowline wax deposition modelling in OLGA, the RRR wax model is preferred to the
Matzain model because it produces a consistent increase of the continuous estimate of the wax
build-up over integration of time for multicomponent mixtures. The trend prediction from the
Matzain model gave a much higher prediction of wax thickness than that predicted by the RRR
model. A simulation verification to confirm the over prediction of the Matzain model was carried
out on the oil with verification result shown by the profile plots of Figure 4.4. The over
prediction of the wax deposit thickness of the Matzain model is attributed to incorporate
deposition enhancing coefficients introduced into the Matzain correlations. This enhancement
coefficient is believed to account for other wax deposition enhancing mechanism, which the
diffusion coefficient and the shear dispersion coefficient could not validate.
As shown in figure 4.4, a high deposit thickness was observed as predicted by the Matzain model
with a value of 3.80957mm at 1198.8m (red line) and 0.870068mm for the RRR model at
Figure 4.4: Thickness of wax layer deposited at wall as predicted by Matzain and RRR
model
Figure 4.5 shows the comparison between fluid temperatures calculated by RRR model with
Matzain model. The outlet temperature for the Matzain model is 40.8302°C and 40.837°C for the
RRR model. The temperature of the fluid keeps on decreasing due to the heat loss to the
surrounding. Although, there is slightly different temperature especially for both models, but it
still proves that the temperature decreasing along the pipeline. The trend shows that the
temperature decreases faster along the length of the pipeline from the inlet to the outlet.
Figure 4.6 is a profile plot of the flow regime. As seen from the figure, no separation was
observed between the flow regimes as predicted by the Matzain and RRR model. The flow
regime is 1 meaning stratified flow. Stratified flow deposit wax crystals on the lower section of
the pipe with a decreasing thickness from the bottom to upward positions. This decreasing
behavior is produced due to higher heat transfer rates in the bottom of the pipeline rather than in
upward points. Stratified flow also present soft deposits at the bottom of the pipe and harder ones
Figure 4.6: profile plot of flow regime for RRR and Matzain after 100days
Figure 4.7 is a profile plot of wax mass precipitation rate. From the figure, it can observe that the
wax mass precitation rate predicted by the Matzain model (red line) was lower than that
predicted by the RRR model (black line) with a value of 0.36082kg/s and 0.346767kg/s for the
RRR and Matzain models respectively. At the outlet of the pipeline, a negative wax mass
Figure 4.8 is a pressure profile plot. As shown in figure, an inlet pressure of approximately
1370.2psia (red line) is required to ensure the continuous flow of the crude oil when the Matzain
model was used to predict the pressure profile, while an inlet pressure of 1339.53psia (black line)
is required to propagate the same fluid when the RRR model was used to predict the pressure
profile. At the outlet of the pipeline, the pressure is 189.106psia for RRR model (black line) and
189.048psia for the Matzain model (red line). Thus, a total pressure drop of 1,181.52psia
occurred along the pipeline for the Matzain model and 1,150.424psia for the RRR model.
Figure 4.8: Pressure profile plot for RRR and Matzain model
Figure 4.9 is a profile plot of the overall heat transfer coefficient. At the pipeline inlet, a high
heat transfer was observed with a value of 15.7841W/m2-K (black line) for the RRR model and
16.7855W/m2-K (red line) for the Matzain model. This high overall heat transfer results in a
large deposit thickness as can be seen in figure 4.4 with a thickness of 3.80957mm for the
Matzain model and 0.870068mm for the RRR model. A rapid decrease in the overall heat
transfer coefficient was observed starting from the pipeline inlet to the outlet of the pipeline
since most of the wax deposition occurred at the pipeline inlet. Thus was as a result of the high
Figure 4.10 is a profile plot of temperature and the wax appearance temperature for the RRR and
Matzain model respectively. As can be seen from the figure, the fluid temperature falls below the
wax appearance temperature right from the inlet of the pipeline with no intersection between the
two plots. The wax appearance temperature predicted by the RRR model was the same with that
predicted by the Matzain model with a value varying between 139.169- 136.534°F from the inlet
5.1 CONCLUSION
The RRR wax model is preferred to the Matzain model because it produces a consistent
increase of the continuous estimate of the wax build-up over integration of time for
multicomponent mixtures.
The trend prediction from the Matzain model gave a much higher prediction of wax
thickness than that predicted by the RRR model. The over prediction of the wax deposit
A high deposit thickness was observed as predicted by the Matzain model with a value of
3.80957mm at 1198.8m (red line) and 0.870068mm for the RRR model at 737.601m
continuous flow of the crude oil when the Matzain model was used to predict the pressure
profile, while an inlet pressure of 1339.53psia (black line) is required to propagate the
same fluid when the RRR model was used to predict the pressure profile.
5.2 RECOMMENDATIONS
To gain a very accurate result, an understanding about the wax deposition phenomena is
required.
Predicting the pigging frequency is very relevant, as it is not enough by only using the
data
needed and performing the simulation, and then getting the result, but a profound
Verification of wax deposition models should be done in order to ensure the selection of
the best model for the prediction and to increase the accuracy of the result.
Appendix 1
Amount
S/N Component
(mol)
1 I8-27 26.06624894
2 I28-31 7.922831727
3 I32-34 4.937792096
4 I35-37 3.392621882
5 I38-40 2.405172965
6 I41-43 1.703757483
7 I44-47 1.172258982
8 I48-53 0.751470794
9 I54-62 0.407582251
10 I63+ 0.119683873
11 N8- 9 9.551134182
12 N10-12 9.551134182
13 N13-15 9.551134182
14 N16-19 9.551134182
15 N20-23 4.650093257
16 N24-25 1.441501731
17 N26-27 0.913609287
18 N28-29 0.64035205
19 N30-31 0.465725617
20 N32-34 0.341683065
21 N35-37 0.247689658
22 N38-41 0.173274933
23 N42-45 0.112459605
24 N46-53 0.061545916
25 N54+ 0.018107165
26 Pr 2.39
27 Ph 2.84