Modeling_ simulation and control of pipeline

Document Sample
Modeling_ simulation and control of pipeline Powered By Docstoc
					         Slug control of production pipeline
                              Tormod Drengstig1 and Sissel Magndal1

This paper presents the results of a simulation study on a production pipeline at one of
Statoils installations in the North Sea. Due to low flow rate, there is a slugging
problem in this pipeline. The aim of the study has been to simulate different control
structures in the multiphase flow simulation tool OLGA in order to suppress slugging.
The main result from this study is that the slugging problem for the investigated
pipeline is completely suppressed using a pipeline inlet control strategy with a simple
PID controller. The controlled variable in all cases is the topside choke valve opening,
and the measurements are pressures at different locations of the pipeline.

1 Introduction
During the last decade, the focus on slug control has increased in the oil industry. The
main reason is that many oil fields are at the end of their lives, and that the ratio of oil,
gas and water changes. Hence, the existing production pipelines are not optimal with
respect to the new compositions, and thereby slugging occurs. There are several kinds
of slugging, including terrain slugging, riser slugging and hydrodynamic slugging.
This study focuses on terrain and riser slugging, see Fig.1.


                       I. Slug formation            II. Slug movement into the separator

                   III. Blowout                        IV. Liquid fallback

                       Figure 1. The buildup and generation of a riser slug.

 Department of Electrical and Computer Engineering, Stavanger University College, School of Science
and Technology, P.O. Box 2557, Ullandhaug, N-4091 Stavanger, Norway. Corresponding author:
The profile of the studied pipeline is given in Fig. 2. The subsea choke is situated at
the pipeline inlet at the wells, whereas the topside choke is situated at the pipeline
outlet on the platform above sea level.

                              Multiphase                             Host
    Satellite                                                      Platform
     Field                    Transport

                                                               Choke                           Test
                                                       PT      Valve


                                                       145 m                    Multiphase
                                                       12” Riser       Heater   Flowmeter
     Pressure                                                                                Production
    Transmitter                                                                               Manifold
                     15 m
     Subsea                          12”
    Template                       Flowline

                                     9 km
Well #1    Well #2
                            Figure 2. Schematic overview of the process.

In the literature there are examples of applications where OLGA is combined with
other software such as Matlab in order to simulate advanced control structures. For
instance, the paper of [1] is partly the basis of ABB‟s SlugCon which is successfully
installed at oil fields in the North Sea [2].

2 Problem statement
Slugging can be identified by measurements of pressure, flow and composition as
shown in Fig. 3. Generally, the flow rate changes dramatically with a frequency
depending on the kind of slugging (hydrodynamic vs. riser vs. terrain).

                       Figure 3. Measurements indicating a slugging problem.
As the slug arrives at the platform, it initiates oscillations downstream the topside
choke and this is unfavourable with respect to separation and operation. Moreover, the
wear and tear of equipment increases as long as the slugging problem exists. Hence,
by controlling the flow arriving at the platform there are several advantages with
respect to prolonging reservoir life, increasing product quality, increasing up-time and
easing operation.

There are in general several means to reduce slugging in pipelines. For instance, by
installing large and expensive slug catchers downstream topside choke, the impact on
topside equipment is reduced. Moreover, by reducing the topside choke valve opening
the slugging problem is often reduced. However, the outcome of this is reduced
production, which is unfavourable. The most efficient and inexpensive mean is in most
cases to apply automatic control of the topside choke based on measurements of
pressure and flow along the pipeline.

3 Modelling
It was decided to use an existing OLGA model of the pipeline and to use the built-in
PID block for controllers. Prior to the simulations, the model was tuned to reproduce
the field measurements of the slugging present in the field. A good match of both
pressure variations, slugging frequency and liquid hold-up was achieved. The
experience with the tuning indicates that the model is sensitive to fluid composition
and choke valve parameters. The model is specified with fixed mass flow and
temperature at the inlet, whereas the outlet is specified with fixed pressure. The
pipeline model consists of 116 pipes. For an introduction to slug generating
mechanisms and how OLGA simulates these phenomena, see [3].

4 Results
Based on the existing instrumentation, i.e. pressure transmitters on pipeline inlet and
outlet as shown in Fig. 2, the following control structures have been simulated; i)
pipeline differential pressure control, ii) pipeline inlet pressure control and iii) pipeline
outlet pressure control [4]. The controlled variable is in all cases the topside choke
valve opening. In addition, a simulation using the riser base pressure (not part of the
existing instrumentation) is carried out. The best performance is obtained with ii)
pipeline inlet pressure control, see Fig 4.


                                                     PT                 Choke

                                                     145 m                       Multiphase
                                                                        Heater   Flowmeter
                                                     12” Riser                                Production
                        15 m
        Subsea                    12”
       Template                 Flowline

                                 9 km
   Well #1    Well #2
                             Figure 4. Pipeline inlet control

The simulation results shown in Figs. 5 and 6 starts with the topside choke controller
in manual and 100% open for 15 hours in order to establish a marginal stable pipeline.
The control parameters during this simulation are tuned manually to Kp = 510-6 and
Ti = 700 sec. The reason why Kp is so low, is that the unit for the „measured‟ pressure
in OLGA is Pascal instead of bar. Moreover, it is assumed negligible dynamics and no
gain in the pressure transmitter, and the transfer function for this unit is therefore hm(s)
=1. Hence, in order to obtain a reasonable gain in the loop transfer function, Kp had to
be of that magnitude. The stroke time for the topside choke is 33 secs. and the sample
time is 10 secs.

                   Figure 5. The pressure downstream subsea choke.
                    Figure 6. The controller output to topside choke.
After 15 hours the controller is set in automatic, and is able to stabilize the pipeline in
approximately 5 hours. The controller set point is 87.5 bar, and the stationary control
output is approximately 82 %. In order to visualize the unstable nature of the pipeline
causing the slugging to occur, the controller is again set in manual at 35 hours with an
output corresponding to the average output when in automatic, i.e. 82%. The result is
that the pressure downstream the subsea choke slowly becomes unstable, though this
seems somewhat surprising since the controller output in the period between 20 and 70
hours seams identical (see Fig. 6). However, the controller output during the period of
automatic control do change, though very little. Fig. 7. shows a magnification of
controller output in the period between 24 and 35 hours.

    Figure 7. Magnified controller output for the period between 25 and 35 hours.

As can be seen, the controller makes only minor adjustments to the choke valve
opening, and this demonstrates the unstable nature of the multiphase transport process.

In order to check the stability of the controller, the liquid flow rate from the reservoir
is increased with 10% from 26.4 kg/s to 29.0 kg/s in one step. The results are
presented in Fig. 8, and as can be seen, the controller stabilizes the pressure
downstream the subsea choke after 4 hours.
                   Figure 8. The pressure downstream subsea choke
                and the controller output after a step in liquid flow rate.

5 Conclusion
This paper presents results from a simulation study where the built-in controller
functionality of the multiphase flow simulator tool OLGA is used on one of Statoils
production pipelines in the North Sea. The results indicate that a simple PID controller
acting on the topside choke valve based on measurements of the pressure downstream
the subsea choke is sufficient to suppress terrain and riser induced slugging in this
pipeline. Moreover, the results indicate that only minor manipulations of the choke
valve are sufficient to stabilize the pipeline. Compared to traditional solutions for slug
depression such as slug catchers, the use of active control on topside choke is
favourable from both economic and reservoir points of view.

6 Acknowledgement
Statoil is acknowledged for the permission to publish these data and for the
cooperation during this work.

7 References
[1] Jansen, B., M. Dalsmo, L. Nøkleberg, K. Havre, V. Kristiansen and P. Lemetayer
   (1999), “Automatic control of unstable gas lifted wells”, SPE Annual Technical
   Conference and Exhibition, Houston, Texas, 3-6 October.
[2] Havre, K., K. O. Stornes and H. Stray (2000), “Taming slug flow in pipelines “,
   ABB Review 4/2000, pp. 55-63
[3] Xu, Z. G. (1997) “Solutions to slugging problems using multiphase simulations”,
   Multiphase Metering, Aberdeen, 12-13 March 1997, 11 pages.
[4] Magndal, Sissel (2001), “Suppresion of terrain and riser slugging in a multiphase
   pipeline - A dynamic simulation study”, M.Sc. thesis, Høgskolen i Stavanger