Improving plant efficiency using predictive maintenance

Marco Tiraboschi and Maurizio De Francesco explain how Emerson collaborated with ENEL, Italy, to introduce predictive maintenance procedures, so optimizing equipment performance and reducing unplanned shutdowns and incidents.

Marco Tiraboschi and Maurizio De Francesco, Emerson Process Management, UK

The power generation industry requires continuous guaranteed production with a minimal risk of disruption to supply. In addition, the demands of the market make it essential that power plants provide maximum efficiency and flexibility at all times. Selecting appropriate maintenance procedures is one of the most critical aspects of power station management.

The Brindisi coal fired power plant in southeast Italy
Click here to enlarge image

Being able to successfully identify and prevent imminent failures produces a smoother operation, an optimized maintenance schedule, and helps avoid major economic and environmental repercussions. Routine machinery health-checks enable mechanical failures to be anticipated while allowing maintenance resources to be directed effectively to where they are needed most.

The end result is an economic optimization of resources and improved overall equipment effectiveness (OEE), a recognised measure of plant effectiveness. The most effective maintenance combines these new predictive technologies with traditional maintenance practices. With this in mind, ENEL, Italy’s largest power company, is implementing predictive maintenance at its power plants.

Brindisi power station

The Brindisi coal fired power plant, ENEL’s 2640 MW flagship facility in southeast Italy, produces between 15 and 17 TWh (billion kWh) annually, and is one of many plants in Italy that is adopting vibration analysis as part of a predictive maintenance policy.

Given the size and complexity of the plant, this new approach to maintenance required a carefully measured introduction. Through a technology consultant, Assistenza Specialistica di Torino (ASP), ENEL collaborated with Emerson Process Management to utilise the CSI 2130 Machinery Health Analyser and Emerson’s AMSà‚®Suite: Machinery Health Manager software.

A specialized vibration analysis team was created, consisting of a machinery expert, an electrical expert and a coordinator. Working with Emerson and ASP, the team members received training and assistance during set-up. Using both internal and external expertise, they identified which machines should be monitored and at what frequency. The training and procedures, together with the experience of the operators, very quickly enabled the team to extend checks to over 450 pieces of rotating equipment.

The adoption of these new technologies, combined with continuous support from Emerson, accounted for major successes as soon as the new maintenance procedures were introduced. Data collected and analyzed by the dual channel CSI 2130 provided a precise ‘picture’ of the machinery’s state of health à‚— enabling the discovery of failures and potential failures that otherwise would have been impossible to spot. This has given the predictive maintenance group more confidence and made the team more proactive.

Methodology and results

The ability of the predictive maintenance technologies to diagnose machinery health problems has significantly reduced the number of operating malfunctions. The understanding of machine condition previously based solely on operator experience has now been replaced by an objective analysis based on measurements taken using innovative technologies. Examining the vibration data identifies the precise type, dimension and possible reason for a failure.

The increased understanding has advantages in terms of the continuous operation of plant assets and better scheduling of maintenance tasks. This translates into greater reliability and reduced costs. Some examples, based on the experience of the Brindisi predictive maintenance group, demonstrate the results realised.

Ljungstràƒ¶m air-gas heaters

In 2006, a predictive study of the Ljungstràƒ¶m air-gas heaters carried out every 45 days, revealed an increase in vibration in the starter motor and the reduction gear in both lines of Unit 1 of the Brindisi plant. The vibration analysis highlighted the presence of an anomaly on the oleodynamic joints. Although this was not significant enough to require an unplanned shutdown or even a reduction in plant capacity, it did need to be addressed before the problem got worse. Further investigation during a planned shutdown confirmed the validity of the analysis. The joints were replaced, at a cost of €5000 ($6208), without any poorly timed interruptions to operation or reductions in load. The economic gain is estimated to be approximately €250 000.

Ash exhausters

Ash exhausters are designed for the transportation of ash using a vacuum from the electrostatic precipitators to the silos. Because they are extremely noisy, they are located in soundproof cabins that are hard to access. The high noise level and the limited space available make it difficult to visually evaluate the state of the machine. The exhausters are now monitored using vibration analysis technologies to provide predictive maintenance information. During periodic monitoring, an abnormal increase in vibration values was noted, indicating vertical and axial movement. Analysis of the highly complex data indicated the presence of a problem with the machine, prompting an immediate shutdown. Subsequent inspection showed that the vibrations were not created because of a fault with the rotary engine, but were due to a problem with its anchorage caused by a fracture in the welded joints at the machinery’s base.

This was a significant problem because the separation from the base could have caused serious damage, not only to the machinery, but also to the connected equipment and pipes.

Performing vibration analysis, using the CSI 2130
Click here to enlarge image

Fixing the problem took about two days and avoided a potential €30 000 worth of damage.

This experience has led to the extension of checks to all exhausters, which will deliver a 90 per cent reduction in failures over time.

Aspirator fans

Aspirator fans are high-powered centrifugal fans necessary for maintaining a vacuum within the boiler. They draw combustion gases toward the DeSox desulphurization plant and on towards the chimney.

Given their importance, the fans are remotely monitored by an online system capable of highlighting when threshold values are exceeded. In addition, they are periodically checked by the predictive maintenance team.These checks indicated an anomalous trend in vibration levels. A modest but continuous increase in rotational vibration, not evident in other similar machinery, implied that a gradual, early stage failure was developing. An inspection during the first significant shutdown confirmed that the fan wheel was unbalanced and a crack was found on one of the blades close to the join with the central supporting plate.

A crack on one of the blades close to the join with the central supporting plate
Click here to enlarge image

Trend analysis enabled the anomaly to be identified while the vibration was still at an acceptable level, thus avoiding a potentially life-threatening situation, serious damage to the machinery and significant economic damage due to lost output.

Coal transportation and bunker feed

The coal transportation and bunker feed system consists of a series of belt conveyors à‚— each one driven by one, two or three 6 KV motors that ensure the supply of coal to the pulverizer bunkers. In view of their importance, the plant’s principal 6 KV motors are carefully monitored by the predictive maintenance team in order to highlight early signs of electrical problems. Checks on the coal feed motors indicated the presence of clear lateral lines around the rotator, implying possible fracture of one of the rotor bars. A replacement motor was fitted and the work was scheduled to ensure the bunkers were not left without coal. Inspection of the damaged motor showed the presence of ten cracks in the rotary ring.

State-of-the-art technology

The experience gained by the predictive maintenance group has enabled it to develop a significant database to enable the easy identification of potential faults. In addition, the group is performing vibration analysis using the CSI 2130 on critical machinery that was previously equipped with continuous online vibration monitoring systems, so as to increase the amount of predictive information available for key plant equipment. This approach will maximise plant safety.

Dismantling the equipment confirmed the presence of longitudinal grooves on the motor’s rotation axle
Click here to enlarge image

Another ongoing activity of the group is to emphasize the potential of predictive maintenance to identify design faults and inadequate technical solutions. This will move the predictive maintenance activities toward proactive maintenance, capable of ‘prompting’ better engineering solutions.

One example of this potential is demonstrated by the modifications made to some flywheel governors (slow turbine rotation motors) that were fitted with roller bearings rather than sliding bearings. When analyzed, the presence of abnormal vibrations clearly showed the ineffectiveness of the existing technical solution. Dismantling the equipment confirmed the presence of longitudinal grooves on the motor’s rotation axle, because it was incompatible with the flywheel governor. It was therefore decided that there was a need to modify the choice of plant machinery.

A health check

The Emerson CSI 2130 is a stand-alone machinery health analyzer that offers data collection, vibration analysis, alignment and balancing in a single unit.

The CSI 2130 is capable of taking measurements of very slow machines (even of just a few rpm) which are out of range for other analyzers. It also works well with faster machines and is capable of measuring signals of up to 80 000 Hz, which enables correct diagnosis on centrifugal compressors and other high-rotation machinery.

The Emerson CSI 2130 standalone machinery health analyzer
Click here to enlarge image

Built-in PeakVue technology detects stress waves à‚— the earliest sign of bearing and gear wear. The PeakVue processing not only offers the earliest warning of developing faults, but also provides an indication of severity. Measurements can be translated into reliable trends to determine the optimal timing for maintenance.

Emerson’s patented dual-channel data collection technique slashes measurement time by up to 60 per cent. This increase in productivity translates into a complete return on investment within the first year.

The unit is small, light and durable and is suitable for field work in a wide variety of industrial applications. Furthermore it has a high level of protection against dust and liquids (IP65), making the CSI 2130 ideal for use in tough environments.

To find out more information on Emerson Process Management, see


The experience ENEL has had at Brindisi clearly demonstrates that extending traditional maintenance by using predictive techniques enables better plant management. It also increases reliability, safety standards and the residual life of machine components, and reduces unnecessary repairs. There has been a major reduction in maintenance costs compared with periodical maintenance or fixed-interval maintenance.

Surveys indicate that current operators have greatly increased their knowledge of the functions of the machines used at the plant and this has made possible the development of more proactive maintenance. The combination of data analysis, experience and reasonable hypotheses forms a winning strategy for failure prediction.

The work undertaken at Brindisi shows that collaboration between a major power company like ENEL and leading asset optimization experts such as Emerson produces excellent maintenance results, maximizing both revenue and plant safety.

No posts to display