You are on page 1of 5

SPE 100308 Case Study: The Application of Inflow Control Devices in the Troll Oil Field

Knut H. Henriksen, SPE, Baker Oil Tools; Eli Iren Gule, SPE, Hydro ASA; Jody Augustine, SPE, Baker Oil Tools

Copyright 2006, Society of Petroleum Engineers This paper was prepared for presentation at the SPE Europec/EAGE Annual Conference and Exhibition held in Vienna, Austria, 1215 June 2006. This paper was selected for presentation by an SPE Program Committee following review of information contained in an abstract submitted by the author(s). Contents of the paper, as presented, have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material, as presented, does not necessarily reflect any position of the Society of Petroleum Engineers, its officers, or members. Papers presented at SPE meetings are subject to publication review by Editorial Committees of the Society of Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O. Box 833836, Richardson, TX 75083-3836 U.S.A., fax 01-972-952-9435.

Although exhibiting a large volume of oil in place, the thin oil column was initially not considered economical for development. However, the construction of increasingly longer horizontal sections (Figure 2), the implementation of multilateral well technology, and a novel sand screen completion have made the field development successful.

Abstract This paper describes the application of Sand Control Screens coupled with Inflow Control Devices (ICDs) in the Troll Oil sub sea field development. Technical and functional description, qualification, computer modeling and production experience verifies that completions with ICDs yield higher volumetric recovery of oil from each well compared with more conventional sand control completion methods. Improved sand control results from the screen systems inflow distribution functionality. Introduction Located in the North Sea 80 km west of the Norwegian West Coast, the Troll Oil sub sea development is one of the Norwegian continental shelfs largest oil producing fields. The thin oil column (4-27m) represented from the start a tremendous challenge, both in terms of drilling and completion operations. Numerous technological as well as operational barriers have been broken1. From installing single 1000 m long horizontal sections, current drilling and completion practices incorporate the construction of down hole drain systems like the Starfish well, covering more than 13500 m of reservoir section through 5 laterals (Figure 1).
Figure 2. Field Map of the Troll Oil Field Infrastructure

One of the main contributions to this was Hydros invention and subsequent development of the ICD technology2. This was incorporated into conventional screen technology aiming at balancing the inflow along the long horizontal sections. Troll Completions and Inflow Control Devices From the start the reservoir sections of the Troll wells were placed horizontally near the oil water contact, normally 0.5 m above3, in order to keep maximum distance to the coning potential of the gas cap. The wells penetrate both the high permeability (1-30 D) C-sands and the low to medium permeability (5-600 mD) M-sands. Sand screens were selected according to the formations sand control requirement. Initially a dual screen pre-pack consisting of a 0.012 gap inner round wire, 20/40 mesh resin coated quartz gravel and a 0.012 gap keystone shaped wire outer jacket, mounted on 6 5/8, 13% chrome base pipe. This was considered an optimal screen design for Troll. However, horizontal reservoir sections approaching 4000 m, increasingly tortuous well paths and the introduction of multi lateral well technology4, led to the requirement for a more robust screen design.

Figure 1. Section

Troll Starfish Well, Covers 13542 m of Reservoir

SPE 100308

During 1999, a stand alone screen selection guide was developed5,6. This incorporated the aspects of mud flow back and sand retention testing. The mechanical design criteria such as burst, collapse, tensile, crossing whipstocks, and metallurgy were also tested. Based on this thorough evaluation and qualification, a shrouded coarse weave premium screen, shown in Figure 3, mounted on 6 5/8 and also 5 base pipes was chosen in year 2000. The 5 screen was used in order to reduce drag during installation in the increasingly longer horizontal sections.

At present Troll has utilized flow resistance elements of 0.2 bar, 0.4 bar, 0.8 bar, 1.6 bar and 3.2 bar for the various well configurations. The flow resistance is given at a flowrate of 26 Sm3/d/ICD. Implementation of ICDs in Reservoir Simulation A method was developed to implement ICDs in the reservoir simulations. Extensive flow tests of ICDs were performed at Hydros Research Centre in Porsgrunn. Rate vs. pressure drop was plotted for each ICD flow resistance rating. Pressure drop curves where generated from these charts.

Figure 3. Premium Screen used in the Troll Field

The ICD flow resistance module, shown in Figure 4, was incorporated into the premium screen design, and increasingly applied in the reservoir completions. A helical flow channel ICD design was chosen.

Figure 6. Pressure Drop vs. Flow Rate Curves for Different Helical Flow Channel ICDs Figure 4. Helical Flow Channel Inflow Control Device

This was subjected to a rigorous qualification testing program, which yielded high accuracy flow resistance profiles at low flow velocities. During the calibration testing, different flow resistance modules where tested at different viscosities. Flow resistance versus flow rate was measured and plotted. The mechanical design of the helical flow channel yielded an accurate flow-resistance compared to the theoretical design calculations. The calibration data is presented in Figure 5.

The multi-segment well model in the reservoir model has been used to represent the wells with ICD in the simulations. The generated lift tables are connected to the segments containing ICDs, one specific table for each segment length of the well. The flow from the reservoir is forced to pass these flow resistance elements prior to entering the wellbore. Hence the correct pressure drop according to the ICD setting is imposed. This method was developed in-house in 1998 and it showed a gain in cumulative oil production by increasing the ICD length. The simulations showed that ICD both gave an increased cumulative oil production, and an increased net present value. Among other things it was clearly demonstrated how the gas break through was delayed with increasing ICD length. This method is now incorporated in the reservoir simulator as the codeword WSEGSICD. The Reservoir Simulation Model Cases The model discussed herein represents a typical Troll well branch, with a 2500 m long horizontal reservoir section. The reservoir is modelled as a 600 m wide and 3000 m long box. The model is dipping 3 degrees perpendicular to the well length. The well is situated 1 m above the OWC in the same sand all along the well. The permeability along the well is 6


W a te r V isc o sity 0 .8 c p


SM= Single lead medium channel TM= Triple lead, medium channel

P re s s u re D ro p ,






0 0 25 50 75 100 125 150 F lo w R a te , b p d SM TM C a lc u la tio n 175 200 225 250 275 300

Figure 5. Calibration Data of Helical Flow Channel ICD

SPE 100308

D. There is a 2 m thick micaseous sand layer with a permeability of 300 mD 1 m below the well, and a 2 m thick M-sand with permeability 300 mD 13 m above the well. The permeability in the rest of the layers is 6 D.
Oil rate, cum. oil and GOR
1600 Oilrate 2500 BASECASE Oilrate ICD-case Cum . oil BASECASE Cum . oil ICD-case GOR BASECASE GOR ICD-case 3000000



1400 2500000 2000 1200

Oil production rate (SM3/DAY)

Gas-oil ratio (SM3/SM3)

2000000 1500

Oil production total (SM3)


Figure 9. Cross Section Along the Well Without ICDs Showing the Remaining Oil at the Last Time Step of the Simulation



1000 1000000


400 500 500000 200



0 01/1998

0 01/2000 01/2002 01/2004 01/2006 01/2008 01/2010 01/2012 01/2014

Figure 7. Oil Rate, GOR and Cumulative Oil Plotted vs. Time for Two Simulations: Base-Case Without ICD and ICD-Case with 3.2 bar ICDs Along the Entire Well

Two simulations are shown. One base-case without ICD and one case with 3.2 bar ICD along the whole well length. The case with ICD gave an increased oil production of 200 000 Sm3 oil in 17 years. The gas break through was delayed by approximately 100 days. After a period with equal growth in GOR, but lower absolute GOR in the ICD well, generally the GOR experience a faster growth in wells with ICD. This is due to the wider spread of the gas cone reaching the well. A more uniform drainage can be observed in the case with ICD. A considerable amount of oil is left in the toe end in the case without ICD. The simulation results can be studied in Figures 8, 9 and 10 below:

Figure 10. Cross Section Along the Well with ICDs Showing the Remaining Oil at the Last Time Step of Simulation

Increasing ICD Coverage in the Troll Wells The ICD setting length has gradually increased from around 30 % of the reservoir section, with stair step setting, to the current practice of 100 % coverage with 3.2 bar all along the well. Continuous completion design optimization based on the evolution of calculation methods and the production experience is the main contributors to this development. The main reason for changing from stair step setting to 3.2 bar only, was to simplify the ICD handling. In this process, the results from the simulations were vital. Optimizing the setting, for the Troll Oil wells, did not give a significant increase in simulated cumulative oil production compared to using 3.2 bar throughout the well. Optimising the ICD setting in a simulation model represents the ideal case. Actual reservoir and well conditions varies from the ideal models, and are difficult to predict. The most significant parameters being annular flow, skin, permeability distribution and contact movements. The reservoir simulator does not take annular flow outside the screen into account, which we know takes place in reality. Having different strengths of ICDs would in addition make the inflow control of the well less efficient due to leakage in the annulus from the strong ICDs towards the less strength ICDs. Accordingly a robust solution comprising an equal setting of 3.2 bar is chosen as best practice for the Troll Oil wells. In addition to the logistical advantage of a fixed ICD strength, the planning has become considerably less time consuming for the completion and production engineers. GOR Monitoring and Management The oil production on Troll is limited by gas processing capacity. It has therefore been made an effort to study the relationship between GOR and liquid production for each



Figure 8. Cross Section Along the Well, Showing the Oil Column at the First Time Step of the Simulation

SPE 100308

Time to GBT, days

individual well7. Through this work it has also been observed that there is a correlation between GOR behaviour and ICD coverage. The longer the ICD, the smoother the well can be operated with respect to GOR control. Contrary to this, the wells with short ICD are very sensitive to changes in choke position. This makes these wells more demanding to operate, and instabilities in the production network can occur. Time to Gas Breakthrough (GBT) The nature of the thin oil rim on the Troll field lends itself to horizontal wells and maximizing the acreage covered by each well in order to reach a large as possible drainable volume of oil. The introduction of more advanced drilling technologies allows steering relative to water oil contact and by this assures an accurate placement of the reservoir sections. Even with the presence of an active aquifer, the predominant reservoir drive mechanism is gas drive. Evident from the Troll West Gas Province extended well test, is that the gas breakthrough will occur almost immediately in a conventional horizontal well completion due to the thin oil layer8. As discussed above, the implementation of ICD in the sand control screen completion distributes the drawdown over the entire horizontal section, hence enabling a balanced inflow profile. This balanced inflow profile promotes an even magnitude of reservoir flux. Hence the gas cone develops evenly. This result in an improved drainage efficiency for the ICD wells compared to conventional horizontal completions. To investigate the effect of increasing ICD coverage on GBT time, all wells of the Troll West Gas Province was plotted in figures 11 and 12. The figures are detailing the individual time to gas break through (GBT) per branch vs. ICD coverage. The dimensionless ICD coverage is represented by the ratio of the ICD length divided by the screen length, for each well.
1200 1000 Time to GBT, days 800 600 400 200 0 0 0,2 0,4 0,6 0,8 1 1,2 ICD-length / screen-length

1400 1200 1000 800 600 400 200 0 0 0,2 0,4 0,6 0,8 1 1,2 ICD-length / screen length

Figure 12. TWGPN: Time to Gas Breakthrough vs. ICD Coverage

As can be seen from the figures above, the data does not offer a conclusive correlation of increasing time to gas break through based on the ICD completion length. This is due to the fact that the data is masked by variables such as: Different points in time of drilling, the oil column thickness decreases while ICD length increases with time Effective well placement relative to the GOC Permeability distribution Distance to no flow boundaries ICD and Available Production Logging Data 8 PLTs have been run in the Troll field, but none in ICD completed wells. One conclusion from the PLT interpretations is that the whole well length is contributing to flow. The inflow profile shows the same signature as the simulated inflow profile in a well without ICD. The North Sea Field Grane has run PLTs in wells completed with ICD. The results confirmed that the method of calculating pressure drop through the ICD, and inflow profile modelling is correct. The data has not yet been published. The Challenge of Annular Flow Comparison testing9 between various screen completion methods, has proved that an ICD screen completion yields near zero velocity annular flow, whereas other open annulus completions yield an annular flow component which represents a focused high velocity flow into the base pipe at the last few feet of the first screen in the heel section. This testing also shows that in a horizontal completion configuration each ICD flow module will take approximately the same rate of fluid into the production string; hence the inflow is evenly distributed over the entire horizontal section. Completion Clean-Up Efficiency Similar to the production phase of the well life, the ICD screen has many positive attributes in the clean up phase. It was realized early on that the Troll ICD wells was cleaned up more efficiently due to the ICD effect compared to non-ICD completions. This has been verified through radioactive tracer technology.

Figure 11. TWGPS: Time to Gas Breakthrough vs. ICD Coverage

SPE 100308

For a 100% ICD screen, closed end completion, the ICD functionality9 will ensure that each screen receives the clean up for the region of the well where it is placed. This is a dramatic reduction in completion mud subjected to each screen, compared to non-ICD screen completions, with an open annulus. Sand Control From a sand control point of view, the ICD screen shown in Figure 13 are required on Troll in order to tolerate the flow back of pre-conditioned completion mud, and subsequent capture of dislodged formation sand of varying grain size distribution, without impairing the production capacity.

Acknowledgements The authors would like to thank Hydro, the Troll field partners and Baker Oil Tools for their pursuit of innovative technology and for allowing the information herein to be printed. References
1. Mikkelsen J.K., Norheim T., and Sagatun S. I. Norsk Hydro ASA, The Troll Story, OTC 17108, 2-5 May 2005. Brekke K. and S.C Lien: Norsk Hydro ASA, New Simple Completion Methods for Horizontal Wells Improve Production Performance in High Permeability Thin Oil Zones. SPE Drilling and Completion September 1994, pp 205-209. T. Madsen and M. Abathi, Norsk Hydro ASA.: Handeling the Thin Oil Zone on Troll., OTC 17109, May 2005. A. Cuthbert, SPE, Sperry/Sun Drilling Services,a Halliburton Company, J. M. Hegreberg and E. Skuglund, Hydro ASA, Stacked Multilateral System for the Troll Olje Field - Development and Case History, SPE/IADC 79866, February 2003. Haaland, A., Rundgren, G., and Johannesen, .: Norsk Hydro ASA.: Completion Technology on Troll Innovation and Simplicity. OTC 17113, 2-5 May 2005. Ballard T., Kageson-Loe N., Mathiesen A. M., Norsk Hydro ASA, The Development and Application of a Method for the Evaluation of Sand Screens. SPE 54745, June 1999. J. Hauge and T. Horn, Hydro ASA, The Challenge of Operating and Maintaining 115 Subsea Wells on the Troll Field. OTC 17111, October 2005 B. T. Haug, Hydro, The Second Long Term Horizontal Well test in Troll: Successfull Production From a 13-m. Oil Column With the Well Completed in the Water Zone SPE 24943, October 1992. K H. Henriksen, , J. Augustine, Ed Wood, Baker Oil Tools Integration of New Open Hole Zonal Isolation Technology Contributes to Improved Reserve Recovery and Revision in Industry Best Practices. SPE 97614, Dec. 2005.




Figure 13. Premium ICD screen


In addition to the rigorous qualification testing discussed above emphasis was put on maintaining a low flow velocity flow regime throughout the screen/ICD system, in order to avoid erosion conditions. The sand control performance of the ICD screen completion is improved compared to conventional screen technology. This is due to the fundamental difference in inflow profile between ICD screen completions and conventional completions. No sand production has been observed in the 150 Troll oil horizontal well sections drilled and completed to date. Conclusions 1. The ICD Technology was invented by Hydro and implemented as a critical component to the reservoir completion on the Troll Oil field development. 2. A premium ICD screen design has been qualified with five different strengths of flow resistance. 3. A method of implementing the ICD into reservoir simulations was developed. 4. The reservoir simulation model shows a gain in oil recovery per well using ICDs. 5. Production management shows smoother GOR development in wells with ICDs. Increasing ICD coverage improves GOR control. 6. PLT experience on Troll does not include ICD wells. 7. Clean up of wells with ICD is improved with increasing ICD length. 8. No sand production has been detected in any of the 150 horizontal reservoir sections drilled and completed in the Troll Oil Field.