Resumption of Submarine Operations Risk Report

RESUMPTION
OF SUBMARINE OPERATIONS
RISK REPORT

 

Prepared for the
SUBSAFE Board
01 November 2004

 

 

TABLE OF CONTENTS

• TABLE OF CONTENTS
• INTRODUCTION
  • Relationship to HMCS CHICOUTIMI Board of Inquiry
  • References
  • HIRA Panel
• PROCESS
  • Risk Review
  • Assessment Standard
  • Hazard Assessment
  • Risk Reduction
• RISK SYNOPSIS
• RISK ANALYSIS
  • Data Passed By The HMCS CHICOUTIMI Board Of Inquiry
  • Assumptions
    • Sequence of Events HMCS CHICOUTIMI
    • Mode of Water Entry Into the Junctions
    • Exposed VP Bulkhead Links
    • Unexposed VP Bulkhead Links
    • Design Issues
  • Risks Assessed
    • VP Bulkhead Links
    • Earth Standard Operating Procedures
    • Casing Safety Harness
    • Emergency Breathing System
    • Non-Rechargeable Firefighting Systems
    • Emergency External Voice Communications
    • Beards
    • Running Opened Up
  • Risks Not Assessed
    • Smoke Inhalation Treatment
    • Atmosphere Monitoring Equipment
  • Additional Risks
    • Ice Rail
    • Emergency Internal Communications
• CONCLUSIONS
• RECOMMENDATION
• RISK CONTROL SUMMARY
• Annex A – Risk Assessment
• ANNEX B - HIRA PANEL HAZARD LOG

INTRODUCTION

1.  In accordance with SUBSAFE Board direction, an analysis was conducted to assess the risk associated with resuming operations in HMCS VICTORIA and HMCS WINDSOR.  Both submarines suspended operations following a severe electrical incident and fire that occurred in HMCS CHICOUTIMI on 05 October 04.

Relationship to HMCS CHICOUTIMI Board of Inquiry

2.  The Chief of Maritime Staff (CMS) has ordered a separate Board of Inquiry (BOI) to examine the events that occurred in HMCS CHICOUTIMI on 05 October 2004.  The purpose of this report is to address safety issues in HMCS VICTORIA and WINDSOR with the object of reducing potential risks to an acceptable level in order to resume operations.

3.  Some of the issues and information considered in the HIRA panel and included in this report will be similar to investigations being conducted by the BOI.  Conclusions and/or deductions made in this report should be considered as preliminary.

4. A variety of statements and evidence considered by this panel will also have been presented to the BOI as evidence.  Consequently, this report is classified Protected B and is not releasable without the express permission of the SUBSAFE Board Chair.

References

5. The following documents were considered during the analysis:

1. Upholder Safety Assessment – Prepared for DPA SSK UPHOLDER IPT by the                          , FNC 7187-34/2628RB Issue 1;
2. 3249-220-3 (DGMEPM) dated 27 October 2004 – Materiel Report – HMCS VICTORIA AND WINDSOR;
3. HMCS CHICOUTIMI Fire - Preliminary Report - Fire Origin and Cause Determination Fire Incident – 5 October 2004 dated 22 October 2004 (releasable only by Board of Inquiry); and
4. 3371-1150-SUBSAFE dated 29 October 2004, Record of Discussion/Decision, 8th SUBSAFE Board.

HIRA Panel

6.  A HIRA panel was held in Ottawa 28-29 October 2004.  This panel was composed of subject matter experts (SME) from key agencies associated with submarine operations, engineering and safety, including:

1. Maritime Staff
   1. Cdr Irvine, MSRMS 5 – Chairman
   2. Cdr Bush, DMMCP 3
   3. LCdr Staples, D MAR STRAT
2. ADM (Mat)
   1. Cdr Hallé, DMCM SUBS
   2. Cdr Crewe, DMSS 3
   3. Cdr Boulet, DMSS 4
   4. LCdr Russell, DMSS 2
   5. Mr Bura, DMSS 3
   6. Mr Demers, DMCM SUBS
3. MARLANT
   1. Cdr Mulholland, CO HMCS WINDSOR
   2. Lt(N) Szczepski, EO HMCS WINDSOR
   3. Mr. LT Taylor, MARLANT FTA
4. MARPAC
   1. Cdr Nicolson, MARPACHQ N32
   2. LCdr Cassivi, CO HMCS VICTORIA
   3. LCdr Gensey, MARPAC FTA
   4. PO1 Vansickle, CELEC HMCS VICTORIA

PROCESS

Risk Review

7.  Prior to the HIRA panel. MSRMS 5 examined the 311 existing hazards catalogued in the              Upholder Safety Assessment.  Due to the specificity of the nine key issues identified by the SUBSAFE Board for review, MSRMS 5 recommended against re-evaluating existing class hazards, which do not adequately cover the scope and chain of hazards that needed to be considered in this case.  The panel agreed with the recommendation.  The panel identified 13 hazards for detailed analysis.

Assessment Standard

8.  The risk matrix contained in CMS Submarine License and Certificate Management System (SLCMS) Version 2.0 (draft), as modified by SUBSAFE Board number 7, was used as the assessment standard.  For the purposes of this Risk Analysis, the modes assessed should be construed as meaning:

1. Harbour – the hazard would affect the submarine’s safety in harbour or the ability of the submarine to leave harbour when planned;
2. Surface – the hazard would affect the submarine’s safety while operating on the surface; and
3. Dived – the hazard would affect the submarine’s safety while operating dived.

9.  The values, bands or interpretations assigned to the risk indices and risk categories are key to understanding the risk analysis.  Risk index values and the risk matrix are contained in Annex A.  Risk definitions have are repeated in Table 1 to ease understanding of the following analysis.

Table 1 - Risk Categories

Hazard Assessment

10.   The HIRA panel discussed the potential hazard, referring to references and seeking additional information where necessary, until consensus was reached on the probability and severity indices.  These index values were entered into the hazard log and the resulting risk calculated for each category.

Risk Reduction

11.   After calculating the risk for each category in each applicable mode [1], the HIRA panel discussed strategies to reduce or mitigate the assessed risk.  These risk control options were listed in each hazard entry.  Once the risk controls and their efficacy had been discussed and consensus reached between panel members, the hazard was analyzed again to determine the residual risk [2].

12.   Risk controls generally try to reduce the probability of a hazard occurring and/or the consequences if the hazard does occur.  The HIRA panel was successful, in most cases, in establishing adequate controls to reduce high risks to acceptable risks.  The statistics for the HIRA panel are shown in Table 2.

Table 2 - Risk Reduction

RISK SYNOPSIS

13.   The 13 hazards assessed by the HIRA panel are presented in Table 3.

Table 3 - Risk Synopsis

RISK ANALYSIS

14.   The complete hazard log is contained in Annex B to this report.

Data Passed By The HMCS CHICOUTIMI Board Of Inquiry

15.   The Board of Inquiry (BOI) passed four pieces of information to the SUSAFE Board Chair.  These were made available to the HIRA panel for consideration.  They were:

1. There was seawater ingress into the submarine;
2. There was an electrical event;
3. There was a fire; and
4. There was a possibility that the red molded electrical insulating material commonly known as “red wacker” did not prevent water absorption.

Assumptions

Sequence of Events HMCS CHICOUTIMI

16.   Using the facts passed from the BOI, anecdotal statements from a variety of reliable sources, information passed by the Damage Assessment and Repair Team (DART), the preliminary report of the lead fire investigator, photos of HMCS WINDSOR and professional experience, the panel debated the sequence of events in HMCS CHICOUTIMI.  It was necessary to deduce this sequence in order to make reasoned assumptions about potential safety hazards in the remainder of the class.

17.  The first assumption, that of the sequence of events in CHICOUTIMI, is:

1. That the submarine was running opened up on the surface;
2. That the submarine took a large amount of seawater down the conning tower that resulted in a layer several inches deep on the Control Room deck, was prevented from draining due to a lack of scuppers on 1 deck, and therefore sloshed around on 1 deck in the rough sea state for a protracted period;
3.  That enough water from the Control Room got into the Commanding Officer’s cabin to submerge, partially or completely, the VP bulkhead links under the Commanding Officer’s bunk (see photo 1);

   
   Photo 1 – VP Bulkhead links Under CO’s Bunk in HMCS WINDSOR

4. That these junctions were submerged, partially or completely, for a protracted period;
5. That the VP earth seen in CHICOUTIMI started at one of the VP bulkhead links under the Commanding Officer’s bunk (See photo 2);

   
   Photo 2 – HMCS CHICOUTIMI Under Commanding Officer's Bunk

6. That, as CHICOUTIMI’s crew was trying to localize the earth, additional earths developed in one or more VP bulkhead links under the Commanding Officer’s bunk including at least one that was the opposite polarity to the first;
7. That the two earths on opposite polarities created a short circuit, either through the ship’s structure or through seawater, which caused arcing of the VP bulkhead links significant enough to melt the deck in the Commanding Officer’s cabin in two places (see photo 2); and
8. That the arcing, which stopped when the main power breakers tripped, contributed to the fires on 1 and 2 deck.

Mode of Water Entry Into the Junctions

18.   The panel had anecdotal evidence from experienced VICTORIA class submariners that there have been previous instances of earths on the VP bulkhead links in VICTORIA, WINDSOR and CORNER BROOK.  Although there are no records, the statements were that all three submarines localized earths to the low-lying VP bulkhead links on the after starboard side of bulkhead 56.  It is from one of these occasions that the current reinsulation specification was developed.  Of note, the earths reportedly occurred on the same set of VP links, which are located near a bilge, which has caused these links to be submerged on occasion, and under a pig’s ear (drain) that frequently overflows (see photo 3).  It was also noted that flushing the connections with distilled water (an SOP procedure for earths) removed the earth on occasion.

Photo 3 - VP bulkhead links on Aft Starboard Side of 56 Bulkhead

19.   The assumption that water penetrated the VP bulkhead links in the CO’s cabin in CHICOUTIMI is reinforced by the previous detection of earths on the VP bulkhead links on the starboard aft side of 56 bulkhead.  Working within that assumption, the panel considered three possible ways for water to enter the joint:

1. A physical rupture in the insulating material that allows water to enter;
2. Water being absorbed by the insulation material and passing through it; or
3. Failure of the             sealing bead allowing water to enter under the insulation by capillary action from the end of the “red wacker” insulation (see figure 1).

Figure 1 - VP Bulkhead Link Original Configuration

20.   There is no way to check whether the insulation on the links under the CO’s bunk in CHICOUTIMI was damaged, since it was either destroyed in the fire or removed as evidence by the BOI.  Given the protected location of the junctions, however, physical damage as assessed as the least likely possibility.

21.   DRDC Halifax has done some preliminary testing of “red wacker” insulating materiel removed from WINDSOR.  After submerging a piece of material in salt water for over 24 hours, there was no appreciable weight gain  [4], which indicates that the material did not appreciably absorb water.

22.   This suggests that an alternative entry mechanism, such as failure of the bead of              sealant is a more probable weak point on the link.

Exposed VP Bulkhead Links

23.   The preceding assumption leads to the deduction that VP bulkhead links, or at least their location in proximity to water, are a weak point in the VP system.  Relevant factors that must be considered are the length of time the junctions are exposed to water before an earth develops and, if the failure mechanism is by capillary action, what type of exposure is required for capillary action to take place (i.e. can spraying water on the insulation cause capillary uptake under the red wacker or must the junction be submerged?).  In the cases at 56 bulkhead, there is anecdotal evidence from experienced VICTORIA class submariners that the links were under a drain that permitted frequent water splashing onto the junction and that in at least one submarine (CORNER BROOK), the links were submerged in a bilge.  Based on the deduced timeline, it is assessed that in the case of CHICOUTIMI the 35 bulkhead links were partially or fully immersed for approximately an hour before the first earth was detected.

24.   The preceding assumptions allow the classification of exposed VP bulkhead links to be defined as those that are positioned in such as way as to allow for submergence or protracted and frequent exposure to water.   In the current configuration of the submarines, these are limited to the VP bulkhead links on the aft side of 35 bulkhead in the Commanding Officer’s cabin and on the starboard aft side of 56 bulkhead.

Unexposed VP Bulkhead Links

25.   Unexposed VP bulkhead links are those that are located in such as position that it is extremely improbable that they could be submerged/immersed and very unlikely that they could be subjected to sprays of water.

Design Issues

26.   If the assumption is accepted that the VP bulkhead links in CHICOUTIMI were the starting point of the electrical incident, which then started the fire, there remains a question as to whether these junctions are a safety hazard in the remaining submarines.  Although all VP bulkhead links in original configuration should theoretically be equally resistant or susceptible to water absorption (discounting any possibility of material degradation over time) this does not allow for the differences in manufacture.

27.   Although there has been no indication that there are any differences in link design between individual submarines, the fact that CHICOUTIMI (ex-UPHOLDER) was built in a different dockyard than the other three submarines may be a relevant factor.  There is no evidence to support this possibility one way or the other, but the Vicolet experience [5] makes this issue a legitimate consideration.

Risks Assessed

28.   In light of the suspected weakness of the insulation of VP bulkhead links, the panel decided to examine major components on the system and not restrict itself simply to bulkhead links.  This approach led to the development of five hazards:

1. Insulation on exposed VP bulkhead links (HazID 001);
2. Insulation on un-reinsulated VP bulkhead links (HazID 002);
3. Battery switchboards (HazID 003);
4. Battery through-deck penetrators (HazID 004); and
5. Battery hatches (HazID 005).

VP Bulkhead Links

29.  HazID 001 concerns exposed VP bulkhead links, as defined in paragraph 24.  Factored into the risk assessment of this hazard is the work currently underway to reinsulate (see figure 2) all exposed VP bulkhead links in VICTORIA and WINDSOR and, eventually, CORNER BROOK.  This reinsulation will protect the links regardless of the actual mechanism by which the links were exposed to water in CHICOUTIMI.  The reinsulation specification was developed some time ago to deal with earths on VP bulkhead links at 56 bulkhead and is considered successful [6] (see photo 4).

Figure 2 - VP Bulkhead Link After Reinsulation

30.   DRDC Halifax is currently testing the repair specification.  A representative insulated joint has been exposed to saltwater immersion.  Initial results after 24 hours of constant immersion is that the insulation materiel does not absorb water (verified by weight measurement) and that the insulating properties remain effective (verified by monitoring the current being passed through the immersed junction.  Long-term exposure tests are continuing.

Photo 4 – Reinsulated VP bulkhead links in HMCS WINDSOR

31.   When considering the hazard, the panel took into consideration a modification already made to the Commanding Officer’s bunk that added an inspection panel. This will allow for periodic visual inspection of the high-risk junctions at 35 bulkhead.  The addition of an access panel alone was considered inadequate to reduce the risk adequately and it was therefore assessed as Undesirable.

32.   In order to reduce the risk further, two risk controls are considered prudent.  The first is a put institute a formal inspection regime.  With no long-term data of the reliability of the joint, a simple visual inspection will permit observation of any gross physical changes of the joint that could indicate a problem.  The addition of a monthly planned maintenance regime was considered practical.  A second safeguard is to issue direction to ensure that VP earths on reinsulated junctions are reported by high precedence OPDEF.  This will ensure that an alert is available that will indicate the failure of a reinsulated junction.  With both measures in place, this residual risk was assessed as Acceptable With Review.

Action 1: Create and issue a monthly Planned Maintenance inspection schedule for inspection of high-risk VP bulkhead links (OPI: DA).

Action 2: Issue guidance that any earth detected on a reinsulated exposed VP junction is to be reported as soon as possible by IMMEDIATE OPDEF (OPI: DA).

33.   HazID 002 concerns the susceptibility of unexposed VP bulkhead links.  These links were previously defined in paragraph 25.  The panel assessed the risk associated with these links as Acceptable With Review.   Their location, which makes extremely improbable that they could be sprayed and virtually impossible that they could be submerged, also makes them very difficult to reach.  Reinsulation would be extremely complicated and time consuming because of the number of interference items that would have to be removed first and, considering the low probability assessed for this hazard, would not significantly reduce the risk rating.  There are no known short-term risk controls that would further mitigate the risk associated with these links.

34.   The DA is beginning a feasibility study to consider physical modifications to the VP system that would remove the bulkhead links and use continuous cables/bus bars through bulkheads, as is the practice with the auxiliary power system.  Anecdotal evidence is that the existing design was developed because cables with the 700 VDC rating required by the VP system are too large to permit standard gland penetrations without compromising the design intent of 35 and 56 bulkheads, which are escape bulkheads.  This modification, if feasible with new technology, would most likely be carried out during a major work period (EDWP or Refit) and would effectively eliminate the risk associated with VP bulkhead links.

35.   HazID 003 concerns the Battery switchboard.  After reviewing the design and discussing the switchboard layout with experienced VICTORIA class submariners, the panel determined that there is a low risk of water ingress into the switchboard.  The height of the exposed contacts inside the switchboard is such that the amount of water needed to reach them would already have caused EOP reactions to begin that would included laying off VP power circuits.  Similarly, the switchboard design includes seals on doors that prevent water ingress from sprays, such as might be used during firefighting.  These seals are inspected for defects on every occasion of opening up the switchboard.  The panel assessed the risk of water ingress into the battery switchboard as Acceptable With Review.

36.   One longer-term initiative that will be progressed by the DA is to seek details of a battery isolation fuse reportedly used in RAN Collins class submarines.  It might be possible to adapt this device for use in the VICTORIA class, which would reduce the severity associated with water in a battery switchboard.

Action 3: Investigate RAN Collins class battery isolation fuse (DA)

37.  HazID 004 concerns possible leaks in battery tank penetrators.  Solid bus bars run from inside the battery tanks through the deck and into the battery switchboards.  A rubber insulating boot molded onto the bar insulates these bars.  The hazard is a failure in the insulating material around the bus bar or a leak at the deck level junction that would allow water to leak into the battery tank.  The insulating material begins inside the switchboard and covers the bus bar down to the deck.  This insulated boot is necked and secured to a steel flange welded into the deck by a steel ring and a series of bolts that are tightened to a specific torque (see photo 5).

Photo 5 – Battery Tank Penetrators in HMCS WINDSOR

38.   The insulating materiel on the bus bar risers may be “red wacker”, although this term seems to have become a generic label for any red electrical insulation.  This fact was not considered pertinent, however, since it was the panel’s assumption that “red wacker” was not the mechanism of failure in CHICOUTIMI.  There is indication that additional sealants have been added to bus bar penetrators in at least one submarine, as evidenced by an insulating material painted over the securing bolts (see photo 6).

Photo 6 – Battery Tank Penetrators With Additional Sealant Applied

39.   The panel noted that battery tanks are subjected to a vacuum test before the battery is installed.  One of the systems covered by this test is the integrity of the battery bus bar penetrators.

40.   An additional safety feature, in terms of exposure to water, is that with the exception of one bus bar penetrator, all are located on a step several inches above deck level (see photo 6).  There was significant discussion about the water found inside the insulation on CHICOUTIMI’s mid point link in the after battery.  One possibility discussed was that the two cups of water found inside the junction box might have come down inside a battery tank penetrator, run along the bus bar and collected in the mid point link, which is at the low point of the system.

41.   The panel considered that it was extremely unlikely that the battery tank penetrators had been the route of water ingress because:

1. The water found in the junction box was fresh water and there are no incidences of fresh water flooding on 2 deck in CHICOUTIMI on record or in corporate memory;
2. Visual examination in CHICOUTIMI revealed that there was no evidence that water was or ever had been inside the insulation of the bus bars on either side of the mid point link; and
3. The amount of oxidation damage to the mid point link indicates that the water had been present for a considerable period of time.

42.   The panel also assessed that the water in the mid point link was not a significant hazard, except from a maintenance perspective, because:

1. There were no earth’s on CHICOUTIMI’s mid point; and
2. CHICOUTIMI’s mid point had recently passed a milliamp leak test.

43.   Based on the design of the battery tank penetrators and discounting them as the possible entry point for water discovered in CHICOUTIMI’s mid point link, the panel assessed the risk associated with leakage into the battery tank penetrators as Acceptable.  Although additional risk controls were not considered necessary, the DA reported that metal kick plates are being installed to reduce the risk of accidental physical damage to bus bar penetrator insulation.

44.   HazID 005 concerns the integrity of the battery hatches.  The securing lugs on VICTORIA class battery hatches are a known weak point in that they are easy to damage.  There is anecdotal evidence from WINDSOR’s previous running period that the battery hatch maintains a watertight seal with up to two damaged lugs.  Each hatch also includes a rubber seal that is inspected on every use.  As a known weak point, hatches are specifically checked and guarded during every EOP evolution where there is a possibility of getting water near the hatch.  The panel assessed the risk as Acceptable With Review.  A long-term resolution of the hatch lug issue is being investigated.

Earth Standard Operating Procedures

45.   Anecdotal evidence has suggested that CHICOUTIMI was experiencing earths and was investigating and localizing them when the electrical incident occurred. As a result, investigation of the Standard Operating Procedure (SOP) for earth faultfinding was examined as HazID 006.  After discussion, it was determined that there are no other prudent actions that could be taken because of the difficulty in determining that the earth is on the VP main power bus and then isolating it.  Taking into account the reinsulation of exposed VP bulkhead links, the panel assessed the risk as Acceptable With Review.

46.   As additional mitigation, there should be an investigation to determine if the battery switchboard earth lamps could be used to check whether the earth is on the main power bus.  The current investigation routine involves de-linking the main motor, a time consuming activity and one that might not be prudent considering the operational scenario at the time.  A second, longer term mitigating activity will be to investigate instrumentation and/or modifications to the VP power system to ease localization and isolation of earths.

Action 4: Investigate SOP to use battery switchboard earth lamps to determine if earth is on VP main power bus (SSTG).

Action 5: Investigate instrumentation and modifications to ease earth localization and isolation of earths on the main power bus (DA).

Casing Safety Harness

47.   The addition of this hazard, HazID 007, was based on an anecdotal report that a casing member was washed overboard from CHICOUTIMI while trying to secure the tow lines, got his safety line caught on the casing and was nearly drowned.  Investigation of the harness and standard casing procedures indicates that there is a low risk of this event happening.  Subsequent reports from MARLANT indicate that the sailor in CHICOUTIMI did not get his safety line caught, but got his foot caught under the lower edge of the casing.  It was assessed that the media report that the RN safety diver cut the safety line was probably accurate, but that the diver cut the line so that the sailor could be recovered into the safety RHIB, not because the line was fouled.

48.   An initial consideration was that the casing safety harness should be fitted with a quick release device, but in light of the report from MARLANT and considering that a quick release device might invalidate the purpose of the harness, the panel did not agree that such a modification was required with any urgency.  It would, however, be prudent to review the design and investigate if better equipment is available.  The panel also reviewed the SOP for manning the casing and the guidance delivered to personnel who work on the casing and agreed that no modifications were necessary.  The panel assessed this hazard as Acceptable With Review.

49.   It would be prudent to review the design of the casing to determine if it is possible to prevent future incidents of this nature.  The safeguards that would be needed would have to be weighed against the potential for noise shorts, increases in laminar flow noise, potential for trapping debris under the casing and effect on anechoic capabilities before any decision is made however.

Emergency Breathing System

50.   Two hazards were considered with respect to the EBS system.  HazID 008 concerns the number of PCL couplings available in the VICTORIA class.  Anecdotal evidence received from CHICOUTIMI was that there were minor problems in mobility during her incidents because of a lack of PCL couplings.  The DA has investigated the capacity of the EBS reducers and, based on anecdotal evidence from CHICOUTIMI, who successfully had the bulk of the crew on EBS for 4-5 hours, has determined that system capacity is adequate.  The DA is providing an additional 40 PCL “Y” connectors to each submarine.  This will allow for “fleeting” and will increase crew mobility while on EBS.

51.   There was considerable discussion on the requirement to pre-rig “Y” adapters in the messes.  Consensus was reached that this practice would likely cause more harm from head strikes on protruding connectors than good it might achieve in reducing plug-in time.  After considering all issues, the panel assessed this risk as Acceptable With Review.

52.   The panel identified a number of additional risk controls, although none that would reduce the assessed risk level.  These controls include:

Action 6: Unit specific training is required to ensure that riders and new-joiners know the location of all EBS PCL couplings (Submarines).

Action 7: A modification to “Y” adapter configuration to allow “Y” adapters to be pre-plugged in messes without raising the risk of head strikes should be investigated (DA).

Action 8: Reflective or luminous markings on all PCL couplings should be ensured (DA).

Action 9: NETE should model the EBS system with additional “Y” connectors (DA).  Individual submarines are encouraged to test the onboard systems to build crew confidence (Submarines).

53.   HazID 009 concerns the risk of not being able to don EBS in time to avoid contaminant (e.g. smoke) inhalation.  Post CWP configuration requires 88 masks, however an EC has recently been approved for VICTORIA, based on sea training recommendations, increasing this number to 104.  The DA will implement this EC in all submarines.  The DA has also added 14 EEBD devices to the existing allotment in each submarine based on a survey done in WINDSOR.  Based on a recommendation from CHICOUTIMI, the VICTORIA practice of pre-rigging 17 EBS masks in watch keeping positions will be adopted across the class.

54.   The panel considered the effect of a pre-rigged hose being ruptured.  The volume of air produced by a ruptured EBS hose was considered small enough that it would not pose a credible risk in feeding a fire nor would it pose a hazard to air capacity.

55.   This risk was rated as Undesirable based on the concern that no amount of training can eliminate the possibility of an individual being unable to don a breathing device before inhaling smoke.

56.   One additional risk control was identified for this hazard, although it will not reduce the assessed risk level.  That is investigating the addition of reflective or luminous markings for all EBS lockers.

Action 10: Investigate reflective or luminous markings for EBS lockers (DA).

Non-Rechargeable Firefighting Systems

57.   Based on a report that CHICOUTIMI expended all non-rechargeable fire fighting appliances during her incidents, the panel raised HazID 010.  The panel noted that no agency, including CHICOUTIMI crew, has raised any issues with respect to the amount of equipment.  In fact, CHICOUTIMI satisfactorily dealt with three fires.  The panel also noted that there are a variety of rechargeable and fitted fire fighting systems that would act as backups to the non-rechargeable appliances.  The panel assessed this risk as Acceptable and concluded that no further risk controls were necessary.

58.   The panel also discussed the issue of CHICOUTIMI’s SFU 90 nozzle, which reportedly fell off the hose when the crew tried to use it.  The DA reported that the defect had been rectified in VICTORIA and WINDSOR.

Emergency External Voice Communications

59.   HazID 011 concerned the amount of emergency voice communications equipment carried in submarines.  The consensus of the panel was that the          portable satellite phone was the most capable device currently available on the market.  Each submarine is fitted with two          phones and each phone has two batteries.  There is a charging routine onboard that ensures each battery is maintained at full charge.  With fully charged batteries, a submarine has     hours of talk time and muchmore in standby.  If this talk time is managed with appropriate communications discipline,     hours should be adequate.  Notwithstanding the efficacy of the          phones, emerging technology should be monitored to identify more capable products.

Action 11: Monitor commercial market for introduction of more capable technology (MCP 3).

60.   There has been a suggestion that the Message Handling System (MHS) Uninterruptible Power Supply (UPS) could be used to recharge          batteries in the event of a power failure.  This will be investigated.

Action 12: Investigate ability to recharge          batteries from MHS UPS (FTA).

61.   At present, the          phones must be used from the bridge, which makes them susceptible to weather damage and also makes it difficult to communicate because of wind noise.  Investigation should be carried out to determine a method of allowing any compatible satellite-capable devices onboard submarines to use the satellite antenna in the communications mast.  This would also provide a capability to use the          phones while dived at periscope depth, which has some safety benefits.

Action 13: Investigate means to use compatible satellite devices with communications mast antenna (DA).

62.   In addition to          phones, each submarine is fitted with several alertment devices, including the Indicator Buoy, Expendable Communications Buoys (ECB) and EPIRB emergency beacons.  Although these systems do not offer a voice communications capability, they will alert shore and SAR authorities that would deploy aircraft to the scene.  Each submarine has several portable VHF radio sets that can be used to communicate with ships or aircraft.

63.   The risk associated with this hazard was assessed as Acceptable.  One risk control that will be implemented, but that will not change the assessment, is to disperse the location of the two phones in the submarine to reduce the possibility of losing them both in a single incident.

Action 14: Locate          phones in different locations (Submarines).

Beards

64.   HazID 012 is based on a report that one of the nine CHICOUTIMI smoke inhalation casualties had a beard.  The concern is that this individual’s beard prevented his EBS mask from achieving a seal.  This is an emotive subject and there is a variety of evidence to argue both for and against beards when considering the seals of protective masks.  In the specific case of CHICOUTIMI, anecdotal evidence including a statement from the affected crewmember reveals:

1. That the one bearded individual that suffered from smoke inhalation took 4-6 breaths of smoke laden air before donning his EBS mask;
2. That of all the crewmembers with beards onboard CHICOUTIMI, only the individual that was delayed in donning his EBS mask suffered smoke inhalation; and
3. That the other eight smoke inhalation casualties did not have beards.

65.   Based on the events in CHICOUTIMI, the panel reached consensus that in terms of safety and submarine fire fighting, beards are not a significant risk in achieving a face seal with any of a submarine’s [7] emergency breathing devices. Consequently, the risk was assessed as Acceptable.  One long-term risk control was considered – to find a hood replacement for the current mask that would render the beard issue moot.

Action 15: Investigate the availability of a hood to replace the current facemask for various breathing fitted breathing systems in submarines (DA).

Running Opened Up

66.   Although not on the original list of issues determined at the SUBSAFE Board, given the assumption made by the HIRA panel that the initiating event was a deluge of water down the conning tower, the panel considered it appropriate to review the hazard of running opened up.  In discussing HazID 013 the panel noted that procedures for running opened up, including cycling the tower to lock personnel in and out, are contained in VICTORIA class SOPs.  After review, the panel considered that the existing SOPs are adequate.

67.   The actual decision to run opened up is dependant on a variety of factors that are not easily codified.  It was observed that the decision to run opened up rests with the Commanding Officer with the exception of one or two EOPs (e.g. communications failure with the bridge).  Even in those cases, the Commanding Officer would be present in the control room before the tower was completely opened up and therefore has the effective power of veto, if necessary.  The consensus of the panel was that guidance to Commanding Officers is adequate and that any other conclusions would have to arise from the BOI findings.  Consequently, this hazard was assessed as Acceptable With Review.

68.   The panel observed that there is very limited drainage from 1 deck in the Control Room.  The consensus was that an effort should be made to determine whether it is possible to fit scuppers to permit faster drainage should water inundate the Control Room.

Action 16: Examine feasibility of fitting scuppers to permit faster drainage from the Control Room (DA).

69.   The panel noted that it is alleged that, in the specific instance of CHICOUTIMI, the Commanding Officer was prohibited from diving earlier in the transit.  Anecdotal information from a credible source noted that the RN SUBOPAUTH did not allocate dived operating areas to the submarine because of                                  in the area CHICOUTIMI was transiting.  Had CHICOUTIMI been able to dive in more sheltered waters, it is unlikely that she would have been in the circumstances that led to taking water down the conning tower.  The panel concluded that if this proves to be the case Canadian SUBOPAUTHs should use their influence to increase the probability that Canadian submarines operating under foreign OPCON always have dived operating areas available to them.

Action 17: Influence foreign controllers to allocated adequate dived operating areas to Canadian submarines when under foreign OPCON (CTG 302.9 & CTG 305.9).

70.   The panel assessed that if actions 16 and 17 were implemented, the risk associated with this hazard, currently assessed as Acceptable With Review, would reduce to Acceptable.

Risks Not Assessed

Smoke Inhalation Treatment

71.   The issues surrounding the medical aspects of diagnosis and treatment of smoke inhalation have been handled separately by the medical community and coordinated by the Director Maritime Health Services (DMHS).  DMHS has stated that the results of his examination concluded that the training, equipment carried and procedures used are adequate.  Both the Physicians Assistants in VICTORIA and WINDSOR have been specifically briefed.  The medical system is continuing to examine and update training, a practice of continuous improvement that is common in health professions following any incident.

72.   The HIRA panel concluded that the members did not have a sufficient level of specialist knowledge to make recommendations beyond those already made by the medical community, therefore the risk associated with the medical aspects of smoke inhalation were not assessed.

Atmosphere Monitoring Equipment

73.   Although this was originally listed as a hazard, the panel found it difficult to state what the hazard actually was, since the original statement from a CHICOUTIMI crewmember was that the issue was the time it took to take readings.  In the case of a fire, atmosphere monitoring commences once smoke has been cleared to determine if the clearance has been adequate.  There is no safety hazard if the procedure takes more time than desired.  It simply means that the crew would have to spend more time in breathing apparatus.

74.   One observation that was noted by WINDSOR was that they had an additional real-time gas analyzer (                    ) that was left onboard following the Atmosphere Quality Study recently performed in WINDSOR by DRDC Toronto.  The consensus of the panel was that this piece of equipment could save some time during smoke clearance procedures and that it should be provided to all submarines.

Action 18: Contact DRDC Toronto OPI (Maj Severs) to seek advice on the real-time gas analyzers for use in submarines then procure as a class fit (DA).

Additional Risks

Ice Rail

75.   It was noted that two sections of CHICOUTIMI’s ice rail were damaged during her ordeal.  One section forward near the capstan, apparently a bridging or expansion piece, was loosened and a similar section just forward of the fin was pulled away.  Both sections appeared to be bolted onto longer runs.  The panel had no specific information or photos, but it was the consensus that it would be prudent for the DA to review the design of the ice rail.

Action 19: Obtain data from CHICOUTIMI, review design of ice rail and modify if necessary (DA).

Emergency Internal Communications

76.   The panel noted that CHICOUTIMI was left with few internal communications or alerting systems, even after reaching Faslane.  CHICOUTIMI crewmembers have reported that the two-way radios carried onboard were invaluable.  The CHICOUTIMI Care and Custody crew, who still use them for internal communications alongside in Faslane, have supported this observation.  WINDSOR and VICTORIA reported that each submarine has a number of commercial           two-way radios, which has been recommended by Sea Training.  The panel recommended that the SSTG take steps to identify the requirement for a formal class fit, possibly using the more powerful device (                                                          ) currently being tested in a field trial by WINDSOR.

Action 20: Raise SOCD for class fit of COTS emergency internal communications devices (SSTG)

CONCLUSIONS

77.   There are no Intolerable risks arising from this risk assessment and only one Undesirable risk.  The single Undesirable risk cannot be mitigated because it allows for the possibility of human error when donning EBS in a contaminated atmosphere.

78.   The most critical risk control identified by this assessment is that of reinsulating the high-risk VP bulkhead links on the aft side of 35 bulkhead and the starboard aft side of 56 bulkhead.  This action has already been completed in VICTORIA and WINDSOR and is will be implemented in CORNER BROOK during her current work period.

RECOMMENDATION

79.   None of the outstanding risk controls are considered serious enough either by themselves or in totality to prevent the safe resumption of operations in VICTORIA and WINDSOR.  Consequently, it is the consensus of the HIRA panel that a recommendation be made to CMS that the current operational pause be rescinded and that VICTORIA and WINDSOR resume their scheduled programs.

RISK CONTROL SUMMARY

80.   This section will amalgamate recommended risk controls into a single list.  A suggested is provided for each control that requires action.

Annex A – Risk Assessment

Probability

Severity

Risk Categories

Acceptable (A) - No further action required

Acceptable With Review (R) - Tolerable, but must be reviewed by relevant accountable authority

Undesirable (U) - Should only be tolerated if risk reduction is impractical

Intolerable (I) - Intolerable, risk must be reduced

Risk Matrix

ANNEX B - HIRA PANEL HAZARD LOG

Annex B - HIRA Panel Hazard Log