DM number: 2866390 EVENT INVESTIGATION REPORT Rev. 01 Date: 19/09/2012 Page 1 of 9 Event Investigation Report Eni Norge Synergy no.2822 Saipem Synergy no.23360 BALLAST INCIDENT SCARABEO 8 4.9.2012 Investigation Report Completion Date: 17.9.2012 DM number: 2866390 EVENT INVESTIGATION REPORT Rev. 01 Date: 19/09/2012 Page 2 of 9 CONTENTS 1. LOCATION 3 2. MANDATE FOR THE INVESTIGATION 3 3. INVESTIGATION TEAM 3 4. METHOD OF INVESTIGATION 4 5. INCIDENT BACKGROUND 4 5.1 5.2 5.3 5.4 5.5 6. DATE AND TIME DEPARTMENT INVOLVED EQUIPMENT INVOLVED KEY PERSONNEL INVOLVED RELEVANT PROCEDURES / DOCUMENTATION SEQUENCE OF EVENTS 6.1 6.2 7. 4 4 4 4 4 4 DESCRIPTION OF INCIDENT NOTIFICATION AND EMERGENCY PREPAREDNESS CONSEQUENCES OF INCIDENT 7.1 7.2 4 5 6 ACTUAL CONSEQUENCE POSSIBLE CONSEQUENCES 6 6 8. STATEMENTS FROM INVOLVED PERSONNEL 6 9. FINDINGS 7 9.1 9.2 9.3 9.3.1 9.3.2 9.4 DIRECT CAUSES UNDERLYING CAUSES EVALUATION OF BARRIERS BARRIERS THAT FAILED BARRIERS WORKING REVIEW OF PREVIOUS INCIDENTS RELATED 10. RECOMMENDATIONS TO PREVENT RECURRENCE 10.1 10.2 7 7 7 7 8 8 8 IMMEDIATE ACTIONS LONG TERM ACTIONS 8 8 11. ATTACHMENTS 9 12. REFERENCES 9 DM number: 2866390 EVENT INVESTIGATION REPORT Rev. 01 Date: 19/09/2012 Page 3 of 9 1. LOCATION Field: Wellbore: 2. PL 533 in the Barents Sea 7220-10-1 Salina MANDATE FOR THE INVESTIGATION A joint mandate from Eni Norge and from Saipem was issued. See Attachment 1. 3. INVESTIGATION TEAM Name Position Signature Leif Sandberg Operational Safety Advisor ENI / investigation team leader Signed electronically Joachim Hvidsten Saipem HSEQ manager Signed electronically Eirk L.Folke-Olsen Saipem K-HVO (coordinating main safety delegate) Signed electronically Helge A. Ellingsund Maritime advisor & Master Mariner / Global Maritime Signed electronically Jarle Nilssen Naval architect & Marine Engineer / Moss Maritime Signed electronically Giacomo Allegra Technical Inspector / Saipem Signed electronically th The investigation team went to the Scarabeo 8 on the 5 of September, and left the rig the following afternoon. During this time, a number of people were interviewed and various documents were obtained. An overview of personnel interviewed is given in Attachment 2. Please note that names have been left out in this overview, which thereby only relates to positions. DM number: 2866390 EVENT INVESTIGATION REPORT Rev. 01 Date: 19/09/2012 Page 4 of 9 4. METHOD OF INVESTIGATION Man – Technology – Organization (MTO) model. 5. INCIDENT BACKGROUND 5.1 DATE AND TIME th September 4 2012 at 14:49 hours. 5.2 DEPARTMENT INVOLVED Maritime department at the Scarabeo 8. 5.3 EQUIPMENT INVOLVED The Ballast Control System. 5.4 KEY PERSONNEL INVOLVED Control Room Operator (COOP) on the day shift Stability Section Leader OIM Eni Drilling Supervisor 5.5 RELEVANT PROCEDURES / DOCUMENTATION Se section 12 References. 5.6 WEATHER CONDITIONS Wind speed: 17 - 21 knots Wind direction SSW Wave: 1,25 – 2,5 meters Roll: 1 degree Pitch: 1,2 degrees Heave: 0,6 meters 6. SEQUENCE OF EVENTS 6.1 DESCRIPTION OF INCIDENT At the day of the incident, drilling operations at the Salina prospect in the Barents Sea were taking place on the rig. The POB was 140, and all technical systems of relevance were intact. As the drilling was approaching the estimated reservoir zone, attention was given to disturb the rig stability as little as possible. This in order to make it easy to read the levels in mud pits etc for the drilling department. This means a.o. to inform the rig floor prior to any ballasting operations. On the bridge, the Control Room Operator (COOP) on duty had been at work on the day shift since 07:00 in the morning. Around 14:40 hours, he was attending to his normal duties when he noticed a movement in the rig, indicating that the rig was tilting in the aft direction. He checked for crane operations, because these can give noticeable changes in the stability condition of the rig. The situation did not change, so the COOP started to operate the Ballast Control System to counter the effect. This was done at 14:49, according to the Scarabeo 8 event log (Reference 12.10). His actions had no effect, and the list continued to develop. The COOP tried several measures, to no effect. As he did not understand the situation, he became more and more stressed. Therefore, he sent for the Stability Section Leader (SSL) and the OIM, which have their offices right next to the bridge. The OIM came immediately, and the COOP informed him about the situation and about his belief that crane operations somehow were involved. The SSL DM number: 2866390 EVENT INVESTIGATION REPORT Rev. 01 Date: 19/09/2012 Page 5 of 9 arrived shortly after, and he and the COOP started to work the Ballast Control System at the panels. The SSL had also problems in understanding why the list continued to develop, which probably was due to the perception that the situation was caused by crane operations. Simultaneously, the Eni Drilling Supervisor (DSV) was having his afternoon meeting with the onshore drilling organization. He noticed the developing list, and eventually aborted the meeting with the reason of an developing emergency situation on the rig. He then went to the bridge, and advised the OIM to check the ballast situation. At about the same time, around 15:08, the OIM activated the General Alarm. At 15:12 hours, the SSL was ordered to activate the “close all valves” function on the Ballast Control System. At this point, the rig had an aft list of 7º, which took around 21 minutes to develop. The situation then stabilized, with no further increase in the list. Thereafter, the situation was brought back to normal stability with no list. NOTE 1: The various clocks on the Scarabeo 8 are not fully synchronized. Therefore, the hours stated in this report may vary within a few minutes. NOTE 2: Both the starboard aft sea chest valve – which is a large diameter valve (ND 300) that opens directly to the sea – and the valve for ballast tank #22S were opened at 14:49 according to the event log. This meant that water was flowing into this 1189 m3 tank due to the gravity induced water pressure. The situation prevailed until the “close all valves” function was activated. 6.2 NOTIFICATION AND EMERGENCY PREPAREDNESS The General Alarm was sounded at 15:08 hours, followed by a PA announcement by the OIM where personnel were ordered to muster at the Temporary Refuge, inside at the Main deck level. At 15:14, the OIM updated the crew on the situation through a new PA announcement. At the same time, Eni Norge nd 2 line Emergency Response Organization was notified through the Duty Officer. Saipems onshore organization was notified at 15:16. At 15:23 the OIM made a new PA announcement, informing that the situation was under control and that the rig was being taken back to an even keel. At 15:29 full POB was accomplished. Thereafter, the OIM went down to the muster area to inform the personnel about the situation and the ongoing actions. At 16:26 the situation was considered fully normalized, and the emergency situation was cancelled. A print of the log in the Emergency Control Centre can be seen in Attachment 3. NOTE 3: After the investigation team had left the Scarabeo 8, we have received information that some of the personnel on board claim to have heard the “abandon rig” alarm and not the “general alarm”. Statements from interviewed personnel and the Emergency Control Centre log indicate that the General alarm was used. NOTE 4: This situation was special in that the Emergency Control Centre was also the scene of the incident. Thereby, the personnel present were as much part of combating the incident as they were managing it. This explains why the first meeting was not held according to the standard, although logs were kept. The investigation team deems this to be a justifiable way of handling this particular situation. NOTE 5: The Joint Rescue Coordination Centre (JRCC) in Bodø did not receive notification from the Scarabeo 8, which is contrary to the established routines on board where the Radio Room has this duty. The JRCC received the nd alert from the Eni 2 line ERO in Stavanger. This seems to be an oversight in the heat of the event, which may be explained by the perceived dramatic situation. This also underlines the importance of the established practice with the nd 2 line ERO also notifying the JRCC. The Scarabeo 8 Radio Room Emergency Response Sheet can be seen in Attachment 4. NOTE 6: The long time needed to establish the POB (21 minutes versus normally around 8 to 12) was because of a significant number of personnel (14) that on their own initiative left the indoor muster station and headed for the life boats forward. The reason for this is that the situation was perceived as very dramatic, and it was seen as against all instincts to go inside the rig in stead of going to the life boats. NOTE 7: The drilling section was informed about the situation, and was ordered by the DSV to prepare for Emergency DM number: 2866390 EVENT INVESTIGATION REPORT Rev. 01 Date: 19/09/2012 Page 6 of 9 Quick Disconnect if the situation so required. This did not become necessary. 7. CONSEQUENCES OF INCIDENT 7.1 ACTUAL CONSEQUENCE No damage to personnel, the environment or the installation occurred due to this incident. 7.2 POSSIBLE CONSEQUENCES If the large aft ballast tank of 1189 m3 had been filled completely – which could have happened if the valves had not been shut in – the rig could have taken a trim/list of 12,2º. This could not have led to capsize and thereby loss of the rig, and thereby this incident is not deemed to have a potential for a major accident. There is a possibility for personal injuries for such a list (12,2º), due to objects coming loose and hitting people. This is however difficult to estimate. Since everyone were mustered and no personnel were in the vicinity of large objects like containers and similar heavy items, the investigation team does not see an imminent danger of fatalities for this incident. Small to medium scale injuries might have occurred, but not fatalities. This indicates a Red 2 categorization for personal injuries (serious lost-time/personal injury) according to “Matrix for categorization of undesired HSE events in Eni Norge”. For the environment, a list of 12,2º would have initiated an emergency disconnect of the well. This would have resulted in release of a limited volume of water based mud, in the range of maximum 75 m³. The composition of the water based mud used at the time of the incident can be seen in Attachment 5, which also gives a calculation of the total volume of potential chemicals to the sea. For this incident this amount is 2,7 m³ of yellow chemicals. According to the above mentioned matrix, this indicates a categorization Green 5 for the potential environmental impact. The maximum possible list in this incident may have resulted in some material damage due to objects coming lose and hitting other objects. This is very difficult to estimate. A conservative estimate is between 10 and 25 million NOK, giving a Yellow 3 categorization. This means that the incident is estimated to have a potential Red 2 categorization within Eni Norge, because of the estimated risk for personnel injuries. 8. STATEMENTS FROM INVOLVED PERSONNEL The COOP on duty had prepared a written statement prior to the arrival of the investigation team. This can be seen in Attachment 6. DM number: 2866390 EVENT INVESTIGATION REPORT Rev. 01 Date: 19/09/2012 Page 7 of 9 9. FINDINGS 9.1 DIRECT CAUSES The direct cause of this incident is that the aft sea chest valve and the valve for ballast tank #22S were opened by the COOP on duty and remained open for a fairly long period. It seems as if the valves were opened in order to compensate for a condition (movement of the rig) which changed due to other causes such as crane operations, and new actions were taken without closing down the first action. 9.2 UNDERLYING CAUSES 9.2.1 The COOP in question was not fully qualified to be alone on the bridge on this duty. Up till one week before the incident, there had been 2 COOPs on duty on the Scarabeo 8. This is due to a Saipem requirement that there shall be 2 COOPs on duty during DP operations and during anchor handling. This requirement has been met by having one Saipem COOP which is fully trained, and one contractor COOP which normally does not have the On the Job Training (OJT) that is mandatory for the Saipem COOPs. When there was no longer a procedural need for 2 COOPs, the Saipem COOP had 14 days on shift and returned to shore. The contractor COOP was asked to take the senior position – meaning being alone on the bridge. This COOP was hired in from an agency. Based upon the interviews, it is unclear to the investigation team why the contractor COOP was put in this senior position, which he had for about one week before the incident occurred. 9.2.2 The system for selecting personnel to safety critical positions on the rig has deficiencies. In spite of the requirement for senior (sole) COOPs to have gone through OJT for this position, a contractor COOP was placed in this position without any compensating measures. It may seem that the contractor COOP was hired to do a junior job, but that the developments led to him being the choice in stead of getting a Saipem COOP mobilized from shore. NOTE 8: On the Scarabeo 8, there have been a total of 6 Saipem COOPs until very recently. 2 of them have been promoted within Saipem, and one has resigned. NOTE 9: The guideline for the Activity Regulations §31 indicates that there shall be at least 2 persons on duty to handle the monitoring and control functions in the central control room on permanently manned facilities. The investigation team has information that indicates that other rig owners in Norway have 2 COOPs on duty on their rigs. 9.2.3 The OJT for the Saipem COOPs on the Scarabeo 8 is not mandatory for contractor COOPs. It is the view of the investigation team that the likelihood for this incident would have been significantly smaller if the COOP on duty at the time of the incident has had this training. 9.2.4 There are indications that the Bridge Handover does not contain a full ballast report. This is a requirement in the Ballast Operations manual section 5.3.4, which indicates 7 issues that shall be addressed in the Bridge Handover. 9.2.5 The man-machine interface on the Ballast Control System and on the bridge in general is not optimal. On the Ballast Control System, it was difficult to identify that the aft sea chest valve was open. Further, the orientation and layout of the screens could be more intuitive. Both these issues added to the stress for the two persons at the Ballast Control Station. Concerning the bridge, it is very large with some 165 m². Especially when alone on duty, it can be very stressful to handle simultaneous operations since one needs to go from station to station to get the overview and reach communication stations. 9.3 EVALUATION OF BARRIERS 9.3.1 BARRIERS THAT FAILED 9.3.1.1 The competence on the bridge was not sufficient in this incident, because the COOP on duty was not fully trained for ballast control tasks on the Scarabeo 8. 9.3.1.2 The process for selecting COOPs for the rig failed, because it was decided to put this contractor COOP into a position for which he was not fully competent. 9.3.1.3 Personnel control was not achieved within an acceptable time, because 14 persons went to the forward life boats in stead of mustering at the place decided by the OIM. 9.3.1.4 Some of the personnel dedicated to the Damage control team did not report to their places, but went directly to DM number: 2866390 EVENT INVESTIGATION REPORT Rev. 01 Date: 19/09/2012 Page 8 of 9 the life boats in stead. 9.3.2 BARRIERS WORKING 9.3.2.1 The intervention by the OIM was decisive in regaining control of the situation. He was also assisted in this by the Eni DSV, which has experience from a similar situation on the Borgsten Dolphin a few years ago. 9.3.2.1 The function “close all valves” on the Ballast Control System worked as intended when it was activated. 9.4 REVIEW OF PREVIOUS INCIDENTS RELATED No similar incidents within Saipem, to the knowledge of the investigation team. One on the Borgsten Dolphin a few years ago as mentioned in 9.3.2.1 above, further details are not known. 10. 10.1 RECOMMENDATIONS TO PREVENT RECURRENCE IMMEDIATE ACTIONS 10.1.1 Immediate implementation of at least one fully qualified COOP on duty at any given time. This means a.o to have completed the OJT in a satisfactory manner. Ref section 9.2.1 above. th 10.1.2 Establish a standing order to close all valves after any ballasting operation. Status: Issued as of September 4 2012, ref SC-8-STO-MAR-010-E. 10.1.3 Issue a reminder to the Stability Section to ensure a proper Bridge Handover for the COOPs. Ref section 9.2.4 above. 10.1.4 Implement requirement for a systematic OJT for all temporary (contracted) personnel in Saipem positions on the Scarabeo 8. Ref section 9.2.3 above. 10.1.5 Issue a reminder to the Radio Room to notify the JRCC on any situations with General Alarm on the rig. This should be a focus also on exercises, preferably with a real call to the JRCC also then. 10.1.6 Set clear and visible requirements for correct mustering in emergency situations, and make sure that this is a topic on safety inductions, training and exercises. 10.2 LONG TERM ACTIONS 10.2.1 Evaluate implementation of 3 COOPs on board the Scarabeo 8 at any given time. Thereby, there can be 2 senior and one trainee COOP where the latter can receive systematic and thorough training. This recommendation is given on the assumption that the Activity Regulation §31 is complied with. Ref a.o section 9.2.1 above. 10.2.2 Evaluate possible measures to improve the man-machine interface on the Ballast Control System. Ref section 9.2.5 above. 10.2.3 Evaluate suitable measures to retain key offshore personnel within Saipem. Ref section 9.2.2 above. 10.2.4 Evaluate if it is possible to install a simulator training module in the Ballast Control station for purposes. training 10.2.5 Evaluate the feasibility of automatic closing of all inlet valves if the heel/trim goes beyond a set value. 10.2.6 Evaluate to have a double acknowledge function and/or an alarm function for critical valves in the Ballast Control System. DM number: 2866390 EVENT INVESTIGATION REPORT Rev. 01 Date: 19/09/2012 Page 9 of 9 10.2.7 Evaluate to change the muster station from within the LQ to at the life boats on the Scarabeo 8. This is a corporate standard in a.o Statoil, and ensures both good control of the mustered personnel and rapid evacuation if this should be necessary. Ref section 6.2 above. 11. ATTACHMENTS Attachment 1: Mandate for the investigation Attachment 2: Overview over interviews during the investigation Attachment 3: Log from the Emergency Control Centre Attachment 4: Scarabeo 8 Radio Room Emergency Response Sheet from the incident Attachment 5: Calculation of potential chemicals to the sea Attachment 6: Written statement from the COOP on duty 12. REFERENCES In this section, references to various reports and procedures are given. These documents are stored in Eni Norge’s electronic document management system DM in the same group at this investigation report and thereby easily available. 12.1 12.2 12.3 12.4 12.5 12.6 12.7 12.8 Filled-in notification form from Eni Norge to the PSA, dated 4.9.2012 SCA 8 Operation Manual - Ballast operations; SCA8-PRO-MAR-016-E SCA 8 Operation Manual - Stability and strength; SCA8-PRO-MAR-017-E SCA 8 Emergency Preparedness Manual; SCA8-MAN-EMG-001-E Scarabeo 8 Manual – Personnel Training Manual; SCA8-MAN-PER-002-E SSDR Scarabeo 8 Training Matrix for Marine Personnel; Revision 2 dated 24.05.2012 Scarabeo 8 Training Manual Control Room Operator: On the job training; SCA8-FORM-PER-013-E Scarabeo 8 Standing order: Actions to be taken by the COOP in situations where there is an unexplained change of heel or trim; SCA8-STO-MAR-010-E 12.9 Extract of CV for the COOP on duty at the time of the incident 12.10 Event log Scarabeo 8 – printed 4.9.2012 at 1626 hours.