HSE RR 195 after 18 years, Pt4 - DP FMEA Solutions?

Introduction:?Limits to the effectiveness of dynamic positioning (DP) failure mode and effect analysis (FMEA) were examined in a UK HSE study in 2004.?It examined the purpose, strengths, weaknesses, and alternatives of DP FMEAs, and recommended improvements in managing the process.?Many of these were implemented and, 18 years later, we still have the same problems.?What was recommended, what was done, and why didn’t it work?

IMCA:?The International Marine Contractors Association was recommended to:

• Produce a DP FMEA Management Guide – IMCA released IMCAM178 a year later and it reflects many of the HSE RR 195 concerns, but all IMCA guidelines are voluntary, must be bought, and can be ignored, unless specified in the contract.?It contains good advice but I don’t think that it provides specific auditable standards or coordinates all stakeholders.
• Update DP FMEA Guidance & Make It More Practical – The study found that work was based on expertise and independent of guidelines.?IMCA produced updated DP guidance.?The volume of guidance increased but the practicality of it was still questionable.?IMCA members resisted change and eventually some operators and clients felt forced to offer stricter competing standards through the MTS DP Committee to improve DP industry safety.?These standards were freely available, rather than pay to read, and came with gap analysis checklists to help ensure application.?DNV-RP-D102 introduced a refinement of my redundancy tables to the industry.?IMCA was forced to catch up.?Guidance has proliferated without necessarily becoming more practical for the FMEA engineer.?There is much more to read but little that practically applies the guidance to the work process.?Much of it explains piecemeal technical concerns to a non-technical audience.?Guidelines should not seek to replace technical training, operating experience, and system engineering from better accredited sources, but need to provide useful principles and reminders.?Guidelines and gap analysis require the user to understand enough about the subject to find gaps in his or someone else’s work, and my own template was too short to link each FMEA subject to strict, clear guidance.?There is a lot of work yet to be performed to close the gap between practice and guidance.?Short of a universal FMEA template and the problems it would create, a better fleshed out gap analysis process could help.?Guidance is voluntary so it needs to be useful.

The problem is still unused guidelines that cannot be directly applied to the work or audited.?Improved guidelines and tools are available but work still depends on the worker.?There is a difference between knowing what’s in the guidelines and effectively using them.?Properly formulated, basic principles generally outperform bloated guidelines but require expert users.

Vessel Operators:?The report did not differentiate between shipyards, vessel owners, vessel operators, and their clients, or recognize the competing desires of each.?Lumped together, they were to:

• Start FMEAs Early – If you want it done right then you need to provide enough lead time to detect and properly resolve conceptual or implementation problems.?This is an economic issue with different parties wanting different approaches.?It tangles up with project management (when is the information available, when does it stop changing, when is it true, & is there any lead time?) and the competing inputs and outputs desired by each participant.?Owners, operators, and clients should have a knowledgeable expert review FMEAs to ensure their needs are covered.??
• Use As-Built Documentation – How many times will the FMEA need done??If you wait until the information is solid then it is too late to fix major errors, but if you review every change then the budget grows with the increased work.?Perfectly good designs have been wrecked by changes made to resolve class comments, and the FMEA needs to reflect the current state of the vessel.?Work based on faulty data is invalidated.?The FMEA should be a living document, but keeping it up-to-date costs money and has some approval risk.
• Use an FMEA Project Plan – Clearly identify objectives, scope, inputs, milestones, etc.?Clearly specify what is needed - not just a class approved DP2 FMEA but one done to particular standards.?What the yard, the owner, the operator, and each potential client wants are not the same.?This is given the consideration that the group paying for the FMEA believes it deserves and we probably need to increase awareness.?As a consultant, I normally need to flesh this out.
• Pay More – If you want it done right then you need to provide enough money to do that.?This is a major issue and is tied to competing project and economic goals.?Low risk is more interested in low cost than high quality, and high risk wants high enough quality at as low a cost as possible.?Who pays and who gets the rewards??What is the point of excellent FMEAs and operation if most clients tend to select a cheaper provider??It’s no wonder that there are problems.?Do not try to compete in some market segments.?Perhaps, the guidelines should recommend minimum DP FMEA price floors for different vessel types to make this issue and the associated problems clearer.
• Involve All Stakeholders – Project managers must love the recommendation to increase interference, oversight, and complication in their busy tasks.?Some of the stakeholders are distant and unknown.?FMEA engineers often don’t have the authority to do this but should provide feedback to their clients and interface with those they are authorized to.?This is done well by some clients.
• Ensure Consultants Follow Good Practice – Yep, it’s in the contract and demonstrated by class approval.?Nope, it’s a problem.?Providers are often selected on reputation, price, availability, and relationship.?Paperwork procedures are often just paperwork, but audit proof.?No provider can offer a meaningful assurance of good work unless it is done by certain expert individuals in the organization but they will often not guarantee that.?Consultant company economics and management do not support good practice due to limited experts, time, and money.?Independent experts and groups focused on high quality market sectors provide the best odds.
• Reflect Actual Operation – Sometimes FMEA engineers make assumptions that make their task easier, and sometimes operators have no intention of paying the extra fuel and maintenance costs required to achieve acceptable redundancy.?Analysis based on false assumptions is not applicable.?A realistic assumption might be poor maintenance, degrading equipment, and occasional compound maloperations, and some market segments’ operations do not reflect good DP practice, but that crosses over into management and crew effectiveness.?The FMEA needs to reflect real operating configurations & loads, and identify redundancy’s technical defenses, without directing management.

A lot of this comes down to the competing economic and management issues that existed before the report.?The study gives good advice but not necessarily advice that will be followed.?The competing interests, participants, and market segments, almost guarantee continued trouble without agreed industry minimums.

Classification Societies:?The study did not recognize the effect of competition between class societies and recommended that they:

• Better Define FMEA Objectives – The study found some mismatch between class societies and with IMO 645.?Class societies have generally improved their DP rules since that time, with some making larger improvements than others.?DNV substantially revised its rules, issued recommended practices, and adopted MTS DP guidelines as part of its improvement.?It seems to be the leader.?ABS has recently adopted some IMCA guidance.
• Ensure Trials Demonstrate Redundancy – The study found little explicit connection between the FMEA and trials, and wanted to tighten the relationship and practicality.?Again, many class societies have improved their requirements, some more than others.
• Develop Common Review Standards – This is a major problem area.?Each class society has their own internal review mechanisms but large vessel owners can get kinder treatment, different offices work at different levels, and the competing class societies offer wildly different levels of enforcement.?This difference may threaten a DP industry leader like DNV, unless their higher standards are preferred by vessel clients and owners.
• Improve DP Rules – Discussed under FMEA Objectives (above).?Some class rules prevent DP FMEAs that only look at simple, benign failures or only at large component groups.
• Encourage Sufficient Time for Review – If the FMEA engineer gets the information late then class gets the FMEA really late and must perform a rush review.?Classification is still focused on main class and does not provide incentives for earlier DP involvement due to competition between the class societies.
• Encourage System Supplier Failure Mode Analysis – Type approval often did not properly cover the failure modes of black box systems and this needed improved.?This is required by some rulesets but is not always done well or visible to FMEA analysts.

Substantial improvements have been made by some class societies while some of their competitors have made little.?Class society standards are under competitive strain, so if the market wants to support quality then it needs to vote with its wallets.

Consultants:?Consultants were recommended to:

• Define & Follow Good Practice – This is an issue that most consulting companies struggle with.?If they are used to working to different standards then the lowest common denominator tends to dominate.?Over time, non-technical, non-experts tend to dominate most companies causing the experts to leave.?Good practice is best supported by consultants led by experts and selecting clients that want the work done right.?The study recommended complying with IMCA guidance unless inappropriate.?Obviously, those behind the MTS DP guidelines found the IMCA guidance lacking.
• Ensure Significant Failure Modes Covered – Both systematic analysis and induction were encouraged.?The most systematic FMEAs reviewed failed to find common mode failures but proper use of the introduced system redundancy tables should make this less likely.?Knowledge of past incidents can be helpful and it is good that IMCA now provides more of this information from behind its pay wall.?While methodology is an aid, performance depends on the next point.
• Provide Competence – You cannot see what you do not know.?FMEA engineers need to have the technical knowledge, operation experience, and system engineering understanding to properly analyze the system.?Most bad reports come from people who do not understand the work and are managed by people who don’t.?Ship and professional engineers have their weaknesses but captains are generally only strong in operational experience.?However, they have strong networks, and tend to dominate the consulting industry.?IMCA only makes things worse as it certifies people without engineering or technical training to perform FMEAs and trials, based on having read the IMCA guidelines and copied a couple documents.?The thought of captains certifying themselves to do FMEAs and trials based on the hints of knowledge found in the guidelines is concerning, but was already too common.?The laws in the US and Canada go to the opposite extreme and require such engineering tasks to be headed by a competent registered professional engineer.?The HSE report noted that FMEAs were most dependent of expert knowledge & practice, and worried about what would happen when the experts were replaced.?The report recommended all FMEAs documenting the expertise of their analysts to make competence more visible.?Few reports do so.
• Ensure System Knowledge – GIGO (garbage in, garbage out).?If the information that the analysis is based on is wrong then the results are invalid and probably wrong.?Late, wrong, and missing information is still a problem, even when the consultants perform careful document control.?I sometimes need to fill large gaps with documentation of identical systems from other projects, clearly identify the assumptions, and verify during trials.?Trials cannot confirm all missing or wrong information.?Sources and their limitations should be documented in each FMEA, but this is often poorly done.
• Communicate Clearly – We generally communicate poorly with insufficient information, excess information, hidden vital information, technical speak, red herrings, poor presentation, and errors.?Shame on us!?The report must be clear, concise, and useful, if it is to be anything but a paperwork exercise.?As the only document that covers some DP risks, this is important to safety.?FMEAs are technical exercises but they need to be readable.?Considerable improvement is still needed.?This might be considered to be in competition with other requirements.
• Link FMEA & Trials – Other types of FMEAs make explicit cross-references between the FMEA failure modes and their test step in the proving trials program to ensure the failure modes are tested and confirmed.?This is not explicit in most DP FMEAs and trials but should be.

Despite the importance to their business, DP companies did not generally follow these recommendations, due to economic, management, and customer pressures.?This will be worsened by retiring experts and certification of non-experts.?Many experts go independent, so they can control these problems.

System Suppliers:?They were to follow quality assurance procedures, provide design philosophy or functional descriptions, consider failure modes early in the design process, and provide failure mode information for external use.?This was already done by some.?It is not clear if things have improved.?More documentation is required but the contents tend towards obfuscation (avoiding problems) and system engineers often don’t know what failures are important outside of their system.?Requiring independent system FMEAs could resolve these problems.?Problems are currently found by test, leaked information, or incident.

Flag States:?The UK HSE was to encourage stakeholders to develop and document good practice.?One flag state that did this was the USCG adopting MTS DP guidelines.

Conclusion:?Substantial improvements have been made over the last 18 years but insufficient improvement to resolve the underlying concerns and contradictions of the market.?There is still room for improvements in guidelines and methodology, but the underlying structural problems from stakeholder misalignment will eventually need to be addressed.?Stakeholders in Canada and the US can take advantage of existing legislation, but some stakeholders would not like that.?So long as people can game the system, it will be gamed to their advantage and disadvantage.

Next Week – Bus Tie – Should It Open, or Stay Closed?

Notes:

John Spouge, author of HSE RR 195, is still an active risk consultant, and can be found on Linkedin

A pdf of the report is available at hse.gov.uk/research/rrpdf/rr195.pdf

The MTS DP Guidelines and tools are freely available at dynamic-positioning.com

IMCA guidelines can be purchased at imca-int.com