A quiet revolution: How computers are changing service and maintenance

Condition based monitoring ensures that servicing takes place only when required Credit: MTU Maintenance Berlin-Brandenburg Turbines


In the last decade, advances in computer processing power combined with availability of fast communications networks have transformed the work of service technicians and engineers, finds Penny Hitchin.

On-site power plants are large, sophisticated bits of kit, involving multiple interconnected components operating around the clock at high temperature and pressure. Keeping them running at optimal performance plays a significant role in operations and yet, on many industrial sites, power production is a sideline for the people in charge of the plant.

Some companies operate and maintain their own power plant, but most outsource the responsibility to specialist contractors who provide the maintenance and servicing needed to keep it in good operational order. Companies providing this service may be a division of the OEM responsible for the manufacture of the turbine, or may be a separate company which specialises in servicing specific models from a single manufacturer

The last decade has seen a transformation in the way that servicing and maintenance is run. The availability of computer processing power combined with the worldwide availability of fast communications networks give service technicians and engineers a much more detailed understanding of the condition of the plant. It also demands that maintenance engineers are now multi-skilled in control and mechanical functions, as well as proficient at interpreting and deploying specialised software.

Careful management can extend a turbine’s life Credit: Wood Group GTS


Regular servicing is the key to efficiency

Gas turbines need servicing at regular intervals. Mike Fisher, president of oil, gas & industrial services for Wood Group GTS, explains, ‘The moment you load up the turbine, the machine starts to degrade. They are on a timer to the end of useful life of that unit.’

Traditional maintenance schedules are based on operational hours. The manufacturers recommend scheduled inspections based on run hours and time periods. A gas turbine might require servicing at 4000 hours or semi-annually. With 8760 run hours in a year, the turbine would have two inspections in a typical year.

But careful management can extend the life of a gas turbine. Fisher says, ‘If you run at 100% load you are absolutely running at the red line of that engine. To extend the life of the unit the best thing you can do is to find that reduction. If you back it off to 95%, the machines will likely run in their comfort zone and degradation rates are much reduced.’

Turbines are designed so that the hot gas path elements of the turbine (combustion liner, nozzles and blades) are operating at their metallurgical limits at these elevated temperatures. However, while a reduction in this firing temperature may improve longevity, reducing load in lower emission engines may be counter-productive as these have to operate at predetermined high temperatures to achieve their design combustion and emission levels.

Condition-based monitoring

Increasingly sophisticated instruments make it possible to measure and monitor numerous factors in the power plant. These data can be used to optimise and control operations and maintenance. Condition-based monitoring can extend service intervals and ensure that intervention takes place only when required.

John Fletcher, general manager of VBR Turbine Services’ UK office, says, ‘We have typical routine inspections, but a lot can happen between routine visits every 4000 hours. You cannot capture sufficient data by visiting the site every six months. Customers want to run their engines as long as possible between service intervals. Condition-based remote monitoring is the way forward.’

Edmund Campion, director of research and development at APR Energy, explains, ‘We use a matrix of manufacturers’ recommendations, best practice based on application performance and historical internal knowledge of what we have seen on the engines.’

Gas turbines suck in enormous volumes of air, and if even a small speck of contaminant (pollution, smoke, dust) gets though the filtration system it will inhibit operations. The first thing that falls off in performance when a turbine struggles to ‘breathe’ is the cooling air to key components. If it loses cooling air, that component burns out.

Engines are set up with instrumentation to look at pressure drops across filtered media. Campion says, ‘We only change air filters, fuel filters and pre-fuel filters when pressures go out of the specification that we have set in agreement with engine manufacturers. That way we don’t damage equipment and get maximum life out of the filter media.’

APR installs power systems across the developing world and fuel quality can be variable. On-site analysis of lube oil and fuel can be beneficial.

Campion says, ‘We have specifications for what we want, but don’t always get it delivered. Sulphur-high fuel is very common. Sulphur in combustion systems produces sulphuric acid which eats away at the inside of the engine and degrades the oil quality at a much higher rate. This has a big impact on service intervals.’

Valuable information from number crunching

High-speed communications enable service companies to use remote monitoring to keep a close eye on what is happening in the plant. Thousands of measurements of temperature, pressure, revolutions, speeds, vibration and timings are captured for analysis.

Wood Group GTS’ Fisher says, ‘We use the remote motoring service to give indications of the health of the machine. Data are streamed to our field service customer base in Houston, Texas and analysed by complex algorithms which alert technicians to any change in status. Monitoring involves looking for exceptions: we study trends and as soon as these change we know there is an issue.’

Trends over time are identified and compared to the plant’s own history and to the operating records of similar plants. This analysis is a recent addition to the service engineer’s toolkit, as developments in technology have increased bandwidths and computing power.

Service companies develop their own proprietary software. VBR Turbines’ latest RAPID (Reliability, Availability and Performance through Intelligent Diagnostics) software is one example. Fletcher of VBR Turbine Partners describes how it is used.

‘We have customers in Russia. Remote monitoring means we can diagnose problems that occur there and actually solve them online from the UK. On-call engineers can dial into the system as soon as they are alerted. It can take as little as two hours from a phone call from the customer to the adjustment of parameters on the system. The old-fashioned way involves getting a call, sorting out a flight and then executing the job, in which case operations are down for at least 24 hours.’

The amount of data collected is increasing. MTU Maintenance Berlin-Brandenburg Turbines’ Uwe Kaltwasser of says that his company typically records 60″70 parameters for gas turbines, although not all of them are continuously transmitted. ‘The more data recorded and observed by remote monitoring systems, the better the indication of engine condition is,’ he says. Other data systems are used to collect information for other parts of the power plant.

Investing in early warning systems

Gas turbine main bearings serve the critical function of supporting and controlling the position of the shaft and rotor. It is not possible to predict damage to bearings but using sophisticated systems can enable early warning signs to be detected from the lubrication oil.

This is another area where technology is advancing. Chip detectors work by absorbing microscopic magnetic particles from the oil, which stick to a sensor that gives a signal, raising an alarm. A sophisticated proprietary on-line lubrication oil debris monitoring system can be installed, providing continuous assessment of the health status of all main bearings in a gas turbine, signalling damage at a very early stage and providing real-time trending information about the development of the damage. This enables operators to quantify the amount of material that has been shed without shutting the gas turbine down, and to swiftly differentiate between a slow wear or startup wear issue and a rapidly developing fault. This can be invaluable in deciding on the appropriate course and speed of action.

Kaltwasser says, ‘On-line oil monitoring systems can give early indication of component failure. This can be linked to the engine control and trigger an early warning. It is an expensive system and customers ask what the benefit is. A bearing failure may result in a significant consequential technical damage of the engine, so it is technology that pays for itself by preventing unforeseen failures.’

His experience is that operators who have experienced such failure will invest in the technology, while others are prepared to take the risk of operating without it.

Siemens plans to deploy laser sintering to replace process nickel to repair gas turbines Credit: Siemens


Online servicing

Increasingly, techniques are being devised to carry out work on a plant while it operates.

The compressor wash is a key part of gas turbine servicing. Dirt accumulates on the air inlet side which compresses air to be burnt. If the machine is offline, a cold wash, using water and detergent, will be used to flush it out. A cold wash involves shutting down the machine for a number of hours. However, an interim PLC-controlled compressor wash can be done online, which increases the interval to the offline wash.

OEMs are developing online hot section borescoping. This process is used for visual inspection work on inaccessible areas of the turbine. A borescope is a tiny camera on a flexible tube suitable for insertion into narrow tubes. Borescoping currently takes place when the machine is shut down. Components are removed from the rear of the engine and a flexible borescope is introduced into the turbine looking for rates of change of deterioration in the hot gas path. Online borescoping has the potential to provide updates on the internal condition of the turbine without disruption to production.

Advances in materials technology

The thermal efficiency of gas turbines can be increased by using components able to withstand higher working temperatures. Single crystal super-alloy turbine blades operate at higher temperatures than crystalline turbine blades. MTU’s Kaltwasser says there is a trend for an increased installation of single crystal material which can be used at increased temperatures and pressures, doubling the life of hot components.

Laser sintering ” similar to 3D printing for plastics ” uses a laser to weld powder particles to create a component. Siemens plans to deploy laser sintering to replace process nickel to repair gas turbine burners. The technique requires a 3D model, powder and sintering equipment. The old tip of the intact burner is turned off and a replacement is printed, reducing repair time from 44 weeks to four. The company says that the first laser sintered SGT-800 gas turbine burners show better material properties than cast ones.

‘Once we are ready for serial production, this will allow us to be closer to the customer and provide better service on demand,’ Siemens’ head developer Vladimir Navrotsky explains.

Power plants need regular attention. Automation can improve efficiency but human interaction will always be necessary.

Wood Group’s Fisher sums up the importance of an interactive servicing regime: ‘All gas turbines are constantly talking to their operators. The question is whether their operators are listening. A gas turbine can fail really quickly so you have to be listening. They are high performing units ” a Formula 1 car, not a tractor ” and they have specific needs. You have to listen to what they are saying.’ This is the role of the multi-skilled service engineer.

Penny Hitchin is a freelance writer, who specialises in energy matters.

More COSPP Articles
Archived COSPP Issues

No posts to display