Ensuring the reliability of sensor data gleaned from condition monitoring can bring dramatic improvements in vessel and equipment performance
The value of noon reports as a basis for analysing vessel performance has been questioned at least since 2013. In that year, a team from University College London noted the dramatic difference in standard error between the traditional, manually transcribed noon report and continuous monitoring systems. Six years later, condition-monitoring experts surveyed by Marine Propulsion still estimate that over 90% of the global merchant fleet use this discredited method of recording vessel performance.
Sjur Clementsen, sales manager for Norwegian vessel-monitoring specialist Kyma, is among those who believe that shipowners have been slow to dispense with the manual noon report. The result is a dramatic increase in the opportunity for error.
“Someone forgets to fill in the report one day and puts down a value similar to yesterday’s,” he says. “That gets sent to shore and someone else punches those numbers into vessel-performance software.”
This is just one way in which modern condition-monitoring practices are stymied by old-fashioned operations. Another is the reliance on inaccurate sensors. Kyma vessel-performance analyst Carlos Gonzalez notes that too often, crew will trust a flowmeter or speed log instead of assessing the reliability of their readings.
Like many performance-monitoring companies, Kyma began as a sensor company, in its case with a popular shaft power meter. That is perhaps why later advances towards broader monitoring and analysis have focused on assuring the reliability of data. A recent example is the Kyma Online notification centre. The new tool is an extension of Kyma’s web-based platform that analyses onboard sensor outputs, reading between 100 and 250 sensors (depending on the vessel) every 15 seconds.
“The focus on reliability will reassure shipowners that may have come to know the adage ‘garbage in, garbage out’ through bitter experience”
Kyma Online can do many things – including benchmarking sister vessels, measuring compliance with charterparty terms and collecting data for regulatory requirements – but before any of this, it must assure data quality.
Mr Gonzalez says: “We have safety measures built in when we take sensor readings. The first is that we set a reasonable range of readings and flag any outliers. Another example is flagging when a sensor is not in consistent communication with the Kyma system. We only use data readings that we know are reliable, and the notification centre makes this transparent.”
The notification centre was launched in August 2018, and has since been deployed for around 100 vessels. The focus on reliability will reassure shipowners that may have come to know the adage ‘garbage in, garbage out’ through bitter experience – including through the time-honoured tradition of the botched noon report.
Improved monitoring of boilers has increased reliability and reduced survey burdens (credit: DNV GL)
French offshore operator Bourbon is also focusing on reliability – first of data, then of its fleet. It has at least two vessels connected to a monitoring system, with more set to join the programme imminently. The goal is ultimately to make better maintenance decisions. But condition monitoring comes first.
“This enables us to ensure that the vessel is operated in an optimal manner, for example for engine power, and to know how precisely,” explains Bourbon head of maintenance Anne-Laure Comte. “Maintenance is then performed according to the use of equipment and not according to a fixed calendar interval.”
The company hopes that this focus on ‘reliability-centred maintenance’ will improve the technical availability of its fleet. Meanwhile, Bourbon is also diving deep into the technical characteristics of components in order to achieve further maintenance improvements. One example is its project with bearings supplier SKF, which involves studying the functions that a piece of equipment must fulfil and assessing the best solution in terms of maintenance.
“Take the specific example of ball bearings,” says Ms Comte. “Systematic change according to the number of hours of use is not necessarily appropriate. On the other hand, a vibration analysis will rapidly identify a change of condition and enable us to remedy it.”
Such investment pays dividends. Classification society DNV GL has reported no avoidable in-service defects across boilers on more than 150 vessels to have adopted its boiler monitoring notation since its launch in 2012. Based on its extensive research into boiler faults, the class society introduced the BMON notation to help shipowners improve boiler performance by reducing deficiencies, downtime and maintenance.
Boosting boiler reliability
DNV GL observed that most defects are caused by corrosion related to water quality, often attributable to insufficient maintenance. Requirements for the notation include regular inspections as well as a prescriptive approach to maintaining the protective lining inside the boiler.
“Lack of a stable and passive magnetite layer on the water/steam side of metal surfaces appears to be the predominant contributory mechanism behind many of the reported defects,” says DNV GL Maritime principal engineer (hull, materials and machinery) Hamid Farahany. Many ship crews also struggle to allocate time and resources – including appropriate materials – to repairs, Mr Farahany adds, exacerbating the situation.
The more rigorous approach to the management of boiler condition in service means that part of the boiler survey can be based on the chief engineer’s inspection report. This arrangement results in greater flexibility over when and where boiler surveys can be carried out, which shipowners have been able to use to reduce vessel downtime in port.
DNV GL says that maintaining the oxide protective layer and monitoring water condition reduces the risk of corrosion in boilers. Avoiding impurities on the heat transfer surfaces also makes heat transfer more efficient, improving fuel consumption and minimising heat stresses.
The notation also requires the optimisation of the boiler’s feed-water system design, including the hotwell temperature, the use of salinometers and the oil content sensor. Combined with monitoring of differential pressures across the exhaust gas boiler, this process minimises the risk of water-side contamination and gas-side defects.
For a cutting-edge example of the benefits that can be derived from the shift to condition-based (or reliability-centred) maintenance, look to the military. Without the commercial constraints of the merchant fleet, and with a demand for absolute reliability, applications in naval vessels often show the extent of what is possible today. In the case of condition-based maintenance, that means using a deep understanding of vessel and equipment condition to extend class surveys indefinitely.
“The condition-based class model solution will help the US Military Sealift Command more efficiently schedule and use resources”
That is the objective of a two-year joint project between the US Navy’s Military Sealift Command (MSC) and classification society American Bureau of Shipping (ABS), which aims to enable the move from calendar-based to condition-based classification. ABS will create digital twins for three MSC vessels and collect data from newly installed hull sensors and from sensors on all classed machinery on board. This will enable the partners to detect abnormal behaviour and provide them with a holistic, real-time view of the vessels’ structural health and performance of equipment.
MSC engineering director Andrew Busk explains that the programme will help the division to achieve a “heightened level of vessel readiness”, supporting timely decisions and enhanced planning of vessel overhaul and repair periods.
“The condition-based class model solution will help MSC target critical areas for repair, prioritise maintenance requirements, and more efficiently schedule and use resources to improve availability,” says ABS chairman Christopher Wiernicki.
Big benefits indeed, made possible by deploying big-data analysis. As shipowners increasingly seek to improve reliability across their fleets, they must maintain a close eye on the reliability of the data they use.