OMAE EFFICIENCY OF ICE MANAGEMENT FOR ARCTIC OFFSHORE OPERATIONS

Transcription

1 Proceedings of the ASME rd International Conference on Ocean, Offshore and Arctic Engineering OMAE2014 June 8-13, 2014, San Francisco, California, USA OMAE EFFICIENCY OF ICE MANAGEMENT FOR ARCTIC OFFSHORE OPERATIONS Badr EL BAKKAY* Total E&P Paris La Défense, France Edmond COCHE Total E&P Paris La Défense, France Kaj RISKA Total E&P Paris La Défense, France * Author is now working at Technip, Paris La Défense, France. ABSTRACT The exploration and production of polar oil and gas fields, which are technologically challenging due to extreme weather conditions, are also constrained by strong environmental issues. Safe and economical activities in such hostile and fragile regions require very insightful engineering. The presence of sea ice is representing a triple challenge: economical, technological and environmental. This makes the Arctic exploration and production activities complex. Ice Management (IM) is one of the tools that could efficiently assist to develop Arctic reserves. However, for each project that uses IM operations, a preliminary study is required to evaluate the efficiency of these support operations and to estimate the possible extension in the season of operation of a field. Efficiency of an IM philosophy can be estimated in a global view based on the extension in the operability window. In a more detailed view, it can be assessed taking into account the optimal number of icebreakers, the IM patterns, the available time for eventual disconnection, and the floe size reduction (leading to ice load reduction). For this study, we will focus on the ice floe size and loads reduction. The most common approach for physical management of sea ice is the one where icebreakers reduce floe size of the drifting ice upstream the floating platform (ref. Moran et al. [1], Coche et al. [2]). This paper describes this philosophy and demonstrates based on real-time simulation that its benefit is limited to mild ice scenarios such as unidirectional ice drift. A more efficient way to manage sea ice is (1) to identify the most hazardous events (e.g. big ridges); (2) prioritize these events; and (3) deal with them starting by the most hazardous one. INTRODUCTION Ice management involves various operational procedures that can be used to reduce ice actions on offshore structure. A complete ice management system includes: ice detection, tracking, forecasting, threat evaluation, ice breaking, iceberg towing and ice alert procedures. In our talk, we will concentrate on the physical part of the ice management. The area is assumed free of iceberg, so only sea ice is considered. For ACEX, ref. [1] and for Shtokman, ref. [2], a typical strategy is presented. This strategy mainly considers two cases: unidirectional drift and ice drift reversal event. This strategy is challenged performing IM Operation simulation. This simplified model concentrates on real ice drift case. Ice drift data being based on ice monitoring in the North Eastern Barents Sea (Nesterov et al. [3]). This model demonstrates the limited efficiency of the above strategy. To improve understanding of the physical ice management need, ice load real-time simulation was performed using an analytical approach. This simulation is concentrating on the ice actions seen by the floating platform. The simulation is performed without any physical management. The aim is to assess when the physical ice management is necessary to support the floating platform (i.e. to reduce ice load to an acceptable level). The findings of both models are discussed in conclusion. ACCRONYMS FP Floating Platform IB Icebreaker IM Ice Management ACEX Arctic Coring Expedition 1 Copyright 2014 by ASME

2 ICE MANAGEMENT CONSIDERATIONS For this study, the ice management patterns are assumed to be circular. This is the conventional approach presented in Moran et al. [1], Coche et al. [3]. The ice management fleet consists of two icebreakers. The primary icebreaker (IB1) which is intended for physical IM operations, she will break the incoming ice into medium floes. The secondary icebreaker (IB2) which will operate downstream of the IB1 and of upstream the platform, she will further break the ice into smaller floes or clear the ice using the azimuth thrusters. In case of unidirectional ice drift, the two icebreakers will be doing circular loops upstream of the platform. During the drift reversal events, they will be doing loops around the platform. Figure 1 and Figure 2 illustrate these IM patterns. Figure 2. Schematic IM patterns in Drift Reversal The IBs are characterized by their performance curve: ship speed in function of ice thickness. In our case, IB1 is able to sail at 2 knots in 1.2 m of ice. For the same thickness, the speed of IB2 is 1 knot. ICE EQUIVALENCY As glacial ice is not uniform over space, the concept of equivalent level ice thickness is used. The equivalent level ice thickness (H ELI ) combines ice parameters and variability in one representative variable. This concept introduced by Keinonen et al. [7] is illustrated in Figure 3. Figure 1. Schematic IM patterns in Unidirectional Drift 2 Copyright 2014 by ASME

3 Figure 4 shows the drift path of this ice floe during the mentioned period. Figure 3. Illustration of Ice Equivalency concept Note that not only different thicknesses are combined in H ELI, but also other ice characteristics such as partial concentrations, total concentration, strength and temperature. From Keinonen [8], [9] and [10], the equivalent thickness for icebreakers speed is given by the equation: Eq. 1. Equivalent Level Ice Thickness C σ : Strength factor C T : Temperature factor C C : Concentration factor Figure 4. Drift path of the ice floe (date from 28/05 to 05/07) (coordinate system WGS84 UTM37) Here we can see that the floe changes its direction often and in an irregular basis. To illustrate the effect of this change in ice drift direction on the efficiency of our IM operations model, we did a simulation on the red square in the figure; a zoom on this square is presented in Figure 5. Where the variable i represents an ice category and P i is the corresponding partial concentration for each ice category, with corresponding thickness and floe size, H i and FS i. H ri is the ridge thickness when applicable. IM OPERATIONS SIMULATION To visualize and understand the approach of the model, especially the loop-method, upstream of (unidirectional ice drift) and around (ice drift reversal) the FP, we elaborated a script to simulate real-time IM operations. Ice drift data for the simulation were taken from an ice expedition technical report. Nesterov [3] presented the results of data acquisition in the North Eastern Barents Sea. This dataset was completed in 2009 as presented in Coche et al. [2]. In the course of the expedition 25 buoys were deployed on ice floes (and 15 on icebergs). The maximum recorded duration of ice floe displacement as the result of the buoys work was days, and the minimum was 1.21 days. Observation frequency was 15 minutes. For our simulation, we chose a buoy that was deployed on an ice floe in point N E on 28/05 at 11:15. This ice floe reached the point N E on 05/07 at 11:00. The Figure 5. Series of severe ice drift reversal events For IM operations (icebreakers loops), we adopted the configuration described previously: Loops upstream of the FP in case of unidirectional ice drift and loops around the FP in case of ice drift reversal. A common and simple approach to distinguish between the two events is to set a speed limit criterion. This is because very slow speeds are associated with ice drift direction change. The approach has been verified from 3 Copyright 2014 by ASME

4 real ice data and presented by Nesterov [3]: speed goes down when ice drift direction changes (heavy scenarios). Figure 7. Simulation of IM operations Real Ice Drift Purple dots: IB1 & IB2, Red rectangle: FP Reversal To conclude, the loop pattern is not satisfactory with a chaotic ice drift pattern. This approach has to be re-assessed. Prior to evaluate alternate patterns, we would like to better understand which events are threats for the FP. Figure 6. Simulation of IM operations Unidirectional Drift - ice equivalent thickness 1.2m A case of unidirectional drift with an equivalent thickness is presented in Figure 6. This case highlights that the loop pattern is efficient. The incoming floes are broken by the primary icebreaker upstream of the FP into small floes, and the secondary icebreaker is breaking them into smaller ones. The ice drift speed was 0.2 m/s. This corresponds to average value of ice drift in [3]. The efficiency of the loop is depending of the ice drift speed, the equivalent ice thickness and the ice capability of the IB. For instance with the same drift speed but with an H ELI of 2m, IB2 is not able to sail upstream the ice! After unidirectional drift simulation, a simulation was run with the ice drift pattern of the red circle in Figure 5. As the drift pattern is chaotic, the IBs are making loops around the FP. The same ice equivalent thickness than in the unidirectional drift case is considered. A series of images are presented in Appendix A. The issue is that in multiple cases, the IBs are downstream of the ice drift compared to the FP position (see Figure 7). In such cases, they are not performing their job (i.e. to break the floes upstream the FP). The FP is thus not protected from ice actions as before. This observation is consistent with Hamilton et al. findings [5]. ICE LOAD REAL-TIME SIMULATION Based on the same ice drift data as for the IM operations simulation, we have done an ice load assessment using a simplified approach taking into account the capability of the FP to change its heading. The assumptions were: No Ice Management operations; 10 tenths ice concentration; Level ice with a thickness of 1.2m; The presence of ridges: a frequency of 7 ridges per km with an average keel depth of 7m; When encountering ridges, the consolidated layer is taken to be equal to 2 times the level ice thickness; FP heading: in level ice, the angle the FP can rotate on its turret in 15min is taken to be 150 (Ref. Bonnemaire et al. [6]); δ is the angle between the ice drift incidence and the heading of the FP (see Figure 8). This angle is used to assess the ice action. 4 Copyright 2014 by ASME

5 5 1 4 cos δ 1 1 sin δ Eq. 2. Ice load on the FP The coefficients of this formula were calibrated on model tests as performed in [6]. This simplified formula permits to assess ice loads during ice drift reversal case and also when the FP is encountering a ridge. The load value is not exact but it permits to assess the severity of the cases encountered. As presented in [4], two thresholds are defined: a target limit at 1000T and an operational limit at 2500T. The operational limit is the threshold for disconnection. Figure 8. Delta angle between the FP heading and the drift direction The FP is assumed to be a ship-shape vessel with icebreaking capabilities as described by Liverov et al. [4]. It means that ice loads are significantly lower if the ice is drifting in front of the platform (δ < 30 ) than if it is drifting sideways. The ice load is assessed using the equation: In Figure 9, five plots are presented. The first one is the H ELI. The peak on the plot is associated to presence of a ridge. The second plot is the angle between the heading of the FP and the ice drift direction. The third and the fourth ones are representing respectively the longitudinal and the perpendicular loads on the FP. The fifth plot is the global load on the platform, calculated using the equation above. On the fifth plot, the target limit and the operational limit are also presented. The ice loads are estimated for the full sequence presented on figure 4 (i.e. 38 days). Figure 9. Real-time load on the FP 5 Copyright 2014 by ASME

6 The complete simulation covers 3647 cases (15-minutes steps). Each case has been classified in below target limit, between target limit and operational limit or above operational limit. The different cases are also sorted in terms of: the presence of ridges or not, ice arriving head-on or transversally, combination of transverse ice and ridges. Table 1 summarizes the results of the simulation. Total (%) (number of cases) Load below Target-limit Load below Operationallimit Load above OP-limit Overall 100% % % % 2 With ridges 100% 95 0% % % 2 With transverse (δ > 30 ) 100% % % % 2 Table 1. Appraisal of the real-time simulation With ridges & transverse 100% 6 0% % % 2 From table 1, it appears that all cases with a ridge provide loads above the target limit. On the opposite 90% of transverse cases give a load below the target limit. That means that managing the ridges is more important than caring about the drift reversals in this configuration. Only 2 cases are above the operational limit. Both cases are associated with ridges and ice drift reversal events. For comparison, the same scenario was performed with different ice thicknesses: 0.7m and 1.5m. For the first ice thickness the operational limit hasn t been exceeded at all, and almost 99% of loads were below the target limit. While for 1.5m ice thickness the percentage of loads above operational limit went up to 0.6% (21 cases out of 3647), and the percentage of loads above target limit exceeded 19% (703 cases). All of these heavy events are associated to the presence of a ridge and/or to a reversal event. DISCUSSION AND CONCLUSIONS From the ice management simulations, we highlighted that classical loop-based ice management operations reduce ice loads in easy ice events such as thin level ice and unidirectional ice drift. But this pattern is not efficient in case of real ice drift with chaotic shape. In function of the ice thickness, the drift speed and the ice capabilities of the icebreakers, the icebreakers are not always upstream of the floating platform to be sheltered. The loop patterns are pleasant as there are simple to explain but they are not adapted to real ice conditions. The use of a simplified ice load models has permitted to identify the case leading to important ice loads on the FP. Due to the icebreaking capabilities of the FP, the threatening cases corresponds to thick equivalent ice (ice ridge, rubble fields, thick ice floes ) associated to drift reversal events. Such cases could be managed by local actions on the thick equivalent ice pieces. To perform smart ice physical management, it is advised first to identify hazardous ice pieces for the FP, to asses those than can interact with the FP and to physically managed them. This approach is more effective than continuous performance of loops around the FP. To conclude, ice conditions should be monitored and potential hazardous events (e.g. big ridges) should be identified and their actual locations communicated to IM vessels. Personal onboard the IM vessels should prioritize these events and proceed to deal with them starting by the most hazardous ones. Once heavy ice broken, IM vessels will be able to go back to their operation zones and keep managing ice near the platform. ACKNOWLEDGMENTS This paper is made possible through the help and support from all Total s Extreme Cold Group members. Special thanks to Dr. Annie AUDIBERT-HAYET for her invaluable advices. REFERENCES [1] Moran, K., Backman, J., Farrell, J.W., 2006, The Arctic coring expedition: drilling at the top of the world. Offshore Technology Conference, May 2006, Huston, paper no. OTC [2] Coche, E., Liferov, P., Metge, M., 2011, Ice and iceberg management plans for Shtokman field. Arctic Technology Conference, February 2011, Huston, paper no. OTC [3] Nesterov A., Sagerup, T., Gudoshnikov Y., Liferov P., Ice drift in the North-Eastern Barents Sea: Measurements and Simulations [4] Pavel Liferov, Guillaume Le Marechal, Marc-Marie Albertini, Michel Metge and Erik ter Brake, Shtokman Development AG, Approach to Performance, Operability and Risk Assessment of the Shtokman Floating Platform in Ice, OTC [5] Hamilton, J., Holub, C., Blunt, J., Mitchell, D., Kokkinis, T., 2011, Ice management for support of arctic floating operations. Arctic Technology Conference, February 2011, Huston, paper no. OTC [6] Bonnemaire, B., Lundamo, T., Serré, N., Fredriksen, A., 2011, Numerical simulations of moored structures in ice: influence of varying ice parameters. 21 st International Conference on Port and Ocean Engineering under Arctic Conditions, July 2011, Montréal, paper no. POAC Copyright 2014 by ASME

7 [7] Keinonen, A., Robbins, I., Icebreaker Performance Models, Seakeeping, Icebreaker Escort. Icebreaker Escort Model User's Guide: Report prepared for Transport Canada, Vol. 3, p. 49. [8] Keinonen, A., Browne, R.P., Revill, C.R. (1989). Icebreaker Design Synthesis, Analysis of Contemporary Icebreaker Performance. Report for Transportation Development Centre of Transport Canada, report TP9992E, Calgary, Canada. [9] Keinonen, A., Browne, R.P., Revill, C.R. (1991) Icebreaker Design Synthesis, phase 2, Analysis of Contemporary Icebreaker Performance. Report for Transportation Development Centre of Transport Canada, report TP10923E, Calgary, Canada. [10] Keinonen, A., Browne, R.P., Revill, C.R., Reynolds, A. (1996); Icebreaker Characteristic Synthesis, for Transportation Development Centre of Transport Canada, report TP12812E, Calgary, Canada. 7 Copyright 2014 by ASME

8 ANNEX A EXTRACTS FROM SIMULATION OF IM OPERATIONS Figure 10. Extracts from simulation of IM operations Real Ice Drift Purple dots: IB1 & IB2, Red rectangle: FP 8 Copyright 2014 by ASME

9 E. Coche / Total EFFICIENCY OF ICE MANAGEMENT FOR ARCTIC OFFSHORE OPERATIONS Badr El Bakkay, Edmond Coche, Kaj Riska, Robert Bridges OMAE

10 INTRODUCTION Titre de la Présentation Lieu et Pays Date Jour Mois Année 2

11 TYPICAL ICE MANAGEMENT PATTERN Unidirectional Drift configuration Drift reversal configuration Titre de la Présentation Lieu et Pays Date Jour Mois Année 3

12 EQUIVALENT ICE THICKNESS C σ : Strength factor C T : Temperature factor C C : Concentration factor Titre de la Présentation Lieu et Pays Date Jour Mois Année 4

13 ICE OPERATION SIMULATOR Ice Management fleet - Scouting icebreaker - Primary icebreaker - Sentinel icebreaker Approach - Continuous circular/elliptical loops Reduction of floe sizes - Time for ramming For ridges - Equivalent thickness - Load assessment Incoming Floe Size Primary IB Sentinel IB Outgoing Floe Size 5 Load

14 IM OPERATIONS SIMULATOR Unidirectional drift Titre de la Présentation Lieu et Pays Date Jour Mois Année 6

15 REAL ICE DRIFT PATTERN Titre de la Présentation Lieu et Pays Date Jour Mois Année 7

16 SIMULATION FOR DRIFT REVERSAL CASE Ice management pattern is not adequate for drift reversal case. Titre de la Présentation Lieu et Pays Date Jour Mois Année 8

17 SIMPLIFIED MODEL WITHOUT ICE MANAGEMENT 2 thresholds - Target limit - Operational limit Disconnection Titre de la Présentation Lieu et Pays Date Jour Mois Année 9

18 ICE ACTIONS ON THE FLOATING PLATFORM Titre de la Présentation Lieu et Pays Date Jour Mois Année 10

19 SUMMARY TABLE Overall With ridges With transverse (δ > 30 ) With ridges & transverse Total (%) (number of cases) 100% % % % 6 Load below Target-limit 96.9% % % 236 0% 0 Load below Operationallimit 3.1% % % % 4 Load above OP-limit 0.06% 2 2.1% 2 0.8% % 2 Ridge => Loads are above Target limit Excess of operational limit is a combination of ridge + reversal case Titre de la Présentation Lieu et Pays Date Jour Mois Année 11

20 CONCLUSIONS Titre de la Présentation Lieu et Pays Date Jour Mois Année 12

21 DISCLAIMER and COPYRIGHT RESERVATION The TOTAL GROUP is defined as TOTAL S.A. and its affiliates and shall include the person and the entity making the presentation. Disclaimer This presentation may include forward-looking statements within the meaning of the Private Securities Litigation Reform Act of 1995 with respect to the financial condition, results of operations, business, strategy and plans of TOTAL GROUP that are subject to risk factors and uncertainties caused by changes in, without limitation, technological development and innovation, supply sources, legal framework, market conditions, political or economic events. TOTAL GROUP does not assume any obligation to update publicly any forwardlooking statement, whether as a result of new information, future events or otherwise. Further information on factors which could affect the company s financial results is provided in documents filed by TOTAL GROUP with the French Autorité des Marchés Financiers and the US Securities and Exchange Commission. Accordingly, no reliance may be placed on the accuracy or correctness of any such statements. Copyright All rights are reserved and all material in this presentation may not be reproduced without the express written permission of the TOTAL GROUP.

22 BACK UP Titre de la Présentation Lieu et Pays Date Jour Mois Année 14

23 IM OPERATIONS REAL-TIME SIMULATION Real ice drift path (Ice expedition 2009, AARI) Unidirectional drift Drift reversal 15 Ice drift reversal events model should be reevaluated - Extrem e Cold Group - Septem

24 ICE LOAD REAL-TIME ASSESSMENT Real ice drift path (Ice expedition 2009, AARI) - FPU heading is taken into account - Uniform ice thickness & 7 random ridges per kilometer - Unique load formula : 38 days of drift 16 Disconnection: combination of ridged and transverse events - Extrem e Cold Group - Septem

25 CONCLUSIONS Development of a preliminary tool: success - Limitations: Shtokman FPU & Barents Sea ice drift speed and floe size distributions - Ridge model: need to be improved and validated by model (or real) tests Real-time simulations - IM operations while unidirectional drift: OK - IM operations while drift reversal: need to be reevaluated Load assessment - Severe cases are a combination of ridged and reversal events Management of severe events in the model 17 - Identification of the most severe objects - Attribute priority levels in function of probability of impact - Extrem e Cold Group - Septem - Manage in priority the severe cases that have a big probability