RPME discusses whether today’s predictive maintenance solutions can deliver on their promises to both boost productivity and prevent accidents on siteFor the Middle East’s refining and petrochemicals sector, predictive maintenance has become an increasingly popular tool for managing operational risks and avoiding production downtime that disrupts commercial and operational productivity.
It involves monitoring often expensive assets at regular intervals to check for any signs of degradation or fault. The idea behind this type of maintenance is to acquire an overall picture of maintenance needs, so that any necessary maintenance work can take place before an equipment failure occurs.
Investing in a system that enables a downstream company to predict both failures and maintenance requirements can prove expensive.
However, minimising downtime and ensuring the smooth running of the production cycle is crucial, particularly as the high costs involved in plant shutdown – when for instance a critical refinery asset fails suddenly – can be astronomical.
“In today’s increasingly complex business environment, predictive maintenance is becoming an absolute necessity, and offers tangible cost saving solutions over routine or time-based preventive maintenance,” explains Steve Hood, general manager at Fluke MEA.
“Predictive maintenance has become a meaningful bridge to prevent avoidable high costs and this is evident in the Middle East oil and gas sector especially as the region’s plants are aging and reserves depleting.”
For the heavy process industry, the cost of downtime in terms of lost productivity and start up yield can be critical, particular as the cost of professional tools and skilled operators far outweigh the cost of forced downtime.
With productivity as the key to survival in today’s competitive global industry, Hood stresses that companies must ensure the reliability of their electrical systems and revenue generating equipment to prevent any costly power failures that can cause malfunction or shut down, and in turn result in excessive energy costs to complete work stoppage.
“Performance trending augments predictive maintenance programmes by documenting operational parameters of facility equipment and processes,” he says.
“This allows plant personnel to establish baselines, track trends, and detect when process and equipment parameters are outside of their optimum ranges.
Article continues on next page …
This data can be used to determine the health or condition of the monitored equipment or processes.” Fluke itself provides a range of devices to implement optimum predictive maintenance, including a range of hand-held electronic test equipment used by electricians, technicians and engineers globally.
As each plant is different, Hood accepts that there is no one-size-fits-all solution. “Instead a blend of the three maintenance methods based on criticality, cost of downtime, age of the plant and profitability is optimum,” he recommends. “One strategy may even be best for one part of the plant and a different one for another.”
According to Bill Broussard, business development manager for asset reliability at Emerson Process Management, predictive maintenance can give downstream companies advanced warning of a particular failure mode coming in order to plan an action and manage risk.
“Certainly predictive maintenance can benefit every asset, but its application on such a grand scale cannot be cost justified,” he points out. For this reason, Emerson’s consultants provide an Asset Priority and a Maintenance Strategy ‘Blueprint’ to drive the proper application of predictive maintenance tools.
“The asset priority is derived by factors, such as safety considerations, regulatory compliance, product quality, process throughput, and operational cost,” explains Broussard. “From there, you map out each asset’s maintenance tasks into a ‘blueprint’ and evaluate where predictive tools can provide valuable time to lower risk of a failure issue.”
The top 30 per cent of assets are good candidates for the application of predictive tools and ultimately both priority and risk-to-fail dictate the maintenance strategy mix of predictive, scheduled, and reactive tasks.
David Hewitt, business development manager for sales at Endress & Hauser Instruments International’s Middle East Support Centre, agrees that predictive maintenance can hold the advantage over both reactive and scheduled maintenance.
“With reactive maintenance, you must wait for a failure, this could lead to an unplanned shut down and loss of production,” he says. “Scheduled maintenance, on the other hand, could lead to servicing assets that do not need any maintenance, generating unnecessary costs.”
In terms of the Middle East, Hewitt is pleased that the region’s hydrocarbon companies are starting to take predictive maintenance more serious, particularly in relation to intelligent instrumentation.
This involves digital communication from field instruments being fed into a company’s Distributed Control System (DCS) to facilitate greater diagnostic information being fed back into the plant asset management systems.
“Intelligent instrumentation enables the operator to understand the process conditions better,” says Hewitt. “If a sensor shows signs of deteriorating, then maintenance can be planned in a predictive manner.”
Article continues on next page …
Predictive maintenance can also play an important role in supporting HSE objectives and avoiding accidents on the plant. “An intelligent device or instrument is always self-checking itself so that any deterioration is noted by the system and an informed decision can be made,” says Hewitt.
“For example, it is now good practice to install an independent and diverse technology on your storage tanks to ensure that no overfilling can take place – intelligent devices offering the SIL 3 category makes this possible.” But for this type of maintenance to work effectively, the decision to use smart instrumentation has to be taken at an early design stage.
“Initially the plant may cost more to build (CAPEX), but operationally (OPEX) there will be cost savings through a more reliable, available and safe plant,” Hewitt says.
Due to the high expense associated with setting up a predictive maintenance solution, some of the region’s oil companies have remained sceptical when it comes to predictive maintenance – with many failing to recognise the long-term gains to overall company productivity.
“The goal is to execute most maintenance in a planned manner and unplanned events cost money,” points out Broussard.
“But the best way to decide where to employ predictive tools on your assets is through understanding the asset’s priority and then to reflect on the maintenance strategy of each asset. If the priority is high, and the failure mode can be managed better with a predictive tool, then employing predictive maintenance tools can be cost justified.”
Emerson itself has seen a significant increase in the sales of predictive maintenance contract services over the past three years.
“The out-source decision has a traditional seven-year cycle typically based upon the economy,” explains Broussard.
“In difficult economic times, companies turn to outsourcing – we have seen the reverse during good economic periods where customer in-source and purchase their own equipment with training.” But ultimately it is the core competencies of any company which dictate the decision to in-source or out-source.
“Predictive maintenance often requires trained resources with specialisations in order to apply the technology effectively,” says Broussard. “Here in the Middle East, many companies are challenged to secure or even train their existing labour resources on these techniques since the competency takes one-to-two years to develop.”
For this case scenario, he recommends that it may be more appropriate to out-source vibration or valve diagnostics and use the information to promote more planned work. For a more comprehensive service contract, the company can provide remote analysis for machinery diagnostics, instruments, and valve diagnostics – all of which provide monthly reporting and analytical feedback.
But for those companies choosing to in-source their predictive maintenance, Emerson also has a suite of predictive technologies on offer for applications on rotating machinery (vibration analysis, oil, ultrasonics, infrared, and alignment/balancing) as well as instrument and valve diagnostics.
By offering companies the choice to select according to their own predictive maintenance needs, the costs of such a system needn’t be prohibitive. As Fluke’s Hood points out, much of the equipment in many facilities already has instrumentation that can be used for performance trending.
“Facilities can install additional sensors if necessary, and easy-to-use and relatively inexpensive data loggers can complete the predictive maintenance performance trending requirements,” he says.
“Tools are available, the tools are affordable, and the tools are easy to use. However, the most effective use of these, or any other quality test equipment, depends on how you apply them to help predict and prevent failures—not just react to them.”
Whether companies decide to out-source, in-source or find a middle ground, the most important criteria for selecting a predictive maintenance solution should be based upon the specific needs of the company in question.
“Assuming you would only out-source to a company that has expertise in predictive maintenance on the specific equipment being maintained, the benefit is that specific expertise,” says Pramesh Maheshwari, general manager EMEA at Honeywell Lifecycle Service Solutions.
“Knowing what to look for is the biggest challenge and while many of the leading indicators of potential problems in complex technology are site specific, knowing what to monitor and all the contributing factors is where the real value is in predictive maintenance.”
In the case of technology such as a distributed control system for example, Maheshwari stresses that it is vitally important to know what to look for as capturing too much information can create a loading problem on the system.
“Where the monitoring has little or no impact and where the conditions you would monitor are universal, the decision to develop that expertise in-house depends very much on the size of the installation, available skills and the mind-set of the company to bring in-house,” he says.
Although the choice of system may depend on a delicate balance between how much a company stands to lose if there is a major equipment failure and the cost of setting up a system, having some kind predictive maintenance in place is a undoubtedly a must.
“Predictive maintenance is the right approach if you are equipped with the right tools to detect right events and trigger points to initiate maintenance in a timely manner and maximise asset utilisation,” says Maheshwari.
“For subsystems without advanced diagnostic tools to monitor equipment performance, it is still best to utilise traditional preventive maintenance techniques to avoid costly downtimes.”
