How Predictive Maintenance of Turbines Improves Power Plant Operations

  Print Print

"Our system detected a vibration step on a 1,000 MW steam turbine unit. By a combination of corrective actions we could avoid an unplanned rotor swap during peak season saving the operator over 20 million euros."

- Kevin Dewaer, Engie Lab

The Challenge:
Delivering remote 24/7 support to power plant operators that experience increased vibrations in their turbine groups.

The Solution:
Developing the Laborelec Vibration Measurement System (LVMS), which is based on CompactRIO, to monitor vibrations, trigger alarms, and provide all necessary data to our central expert center regardless of power plant location, so our experts can give immediate advice on the continued operation or predictive maintenance of the turbine.

Kevin Dewaer - Engie Lab

Laborelec, GDF Suez, and Engie Lab

Engie Lab is a technical competence center in electrical power and energy technology. Until 2016 the company was known as Laborelec, part of GDF Suez, and is now renamed Engie Lab. Engie is the fourth largest electricity producer in Europe and has power plants around the globe. Engie Lab, an NI Alliance Partner, has more than 50 years of experience in independent contract research, services, and operational assistance related to power plants. Within Engie Lab, we have a department specifically focused on vibrations and structural mechanics that developed solutions and services for vibration monitoring of large turbo groups.


What happened with a turbine just before, during, and after the alarm sounded, or when the machine tripped? The Laborelec Vibration Monitoring System (LVMS) can tell you. The LVMS project started 30 years ago, and since then we have added new features and improvements based on the needs of our vibration experts and operation staff. It has evolved into a compact and robust system that users can easily install and maintain. Currently, these systems monitor more than 90 machines worldwide.

All turbines in a power plant come equipped with protection systems to prevent equipment from major damages. These systems apply instrumentation to the turbine and connect to the plant distributed control system (DCS) to deliver status updates. However, for predictive maintenance and turbine monitoring we need to acquire data faster.

Our LVMS system uses the measurement signals of existing sensors and makes it possible to integrate additional sensors. Typically the system acquires dynamic signals coming from relative vibration sensors (proximity probes), absolute vibration sensors (accelerometers or velocity meters), or tachometer pulses, and can be combined with a variety of ancillary sensors. Using this setup, we can make a precise determination of the vibration behavior of the shaft and bearing and calculate the shaft position.

Figure 1. Users can install LVMS, our solution for vibration monitoring of turbo groups, in a cabinet close to the protection system, or they can use a portable system for on-site diagnosis.


The LVMS setup features a CompactRIO system connected directly to the buffered outputs of the protection system and additional sensors we installed. A LabVIEW application in the CompactRIO system acquires the data at a high sampling rate. The application then resamples to 16 samples per rotation, independent of rotational speed, and calculates critical parameters including amplitude, RMS values, harmonics (orders), phases of each harmonics, DC offset as well as power generated, and more.

The CompactRIO system acquires data all the time, including run up, coast down, and steady state. The system also obtains the current operating state of the turbine (for example, peak power delivered). A computer application triggers an alarm in case of abnormal vibration behavior and stores extra data from a set time before and after the event for further analysis. The data is continuously captured and stored to an archive on-site that can hold over a year’s worth of data, which we can use for trending. The data is also transferred to Engie Lab headquarters, where we can view the data and advise on corrective actions.

Figure 2. Overview of the LVMS is shown.

Benefits of the Fourth Generation of LVMS
We developed LVMS 30 years ago on an HP platform. Over time we improved the system by adding sensors, custom electronics, and software features. This resulted in LVMS3, which we based on custom hardware in a 19” rack. The rack contained analog I/O modules to do the signal conditioning for two channels and could accommodate up to eight boards. We connected sensors to the system through a set of terminal blocks connected to the 19” rack. We connected this rack to an NI data acquisition board installed in an industrial computer.

We based this fourth generation of our vibration measurement system on CompactRIO, which empowers us to hook up all the sensors directly to the C Series I/O modules using standard BNC cabling. This eliminates a lot of wiring and noise sources. The CompactRIO now uses only 30 percent of the volume and has 50 percent more channel I/O compared to the previous solution. Data processing happens in our CompactRIO application and we only need a PC for data storage. An additional benefit is NI’s global presence. We have customers around the globe. We can ship spare parts to a customer in Thailand or Chile much easier with NI, giving our customers shorter response time and more availability. Finally, the quality control and checks of the electronics are well maintained within NI and were not our core competency.

Diagnostic Software 

We store acquired data at the customer site and periodically download it to servers in our head office. Over the years we have developed an extensive offline software solution and analysis library to quickly analyze and compare vibration signatures from a variety of turbine types and failure modes, which can be quickly pulled into the application and enhances our diagnosis in case an alarm is obtained.

Figure 3. Our vibration experts can diagnose turbine problems remotely.

Benefits to Operators 

With LVMS, operators have a reliable fault detection system and can call our experts in case of alarms at any time of day or night. The experts have fast (remote) access to all the necessary information, including the historical data to make a good analysis. For example, when an increasing vibration level is detected, we can log in and check to understand the issue and inform the operator if the machine can safely advance to the next level or not.

Figure 4. The LVMS predictive maintenance tool helps to quickly identify the real cause of vibrations and keep our customers’ machines up and running.

This can prevent a machine trip (safety shutdown), potentially saving tens of thousands of euros. In one case we detected a vibration step on a 1,000 MW steam turbine unit. Our root cause analysis showed an important contributor to be an imbalance in the exciter rotor. By a combination of corrective actions, such as performing field balancing and placing dynamic vibration absorbers during a planned short stop, we could avoid an unplanned rotor swap during peak season and prepare a spare rotor for the upcoming planned overhaul. This saved the operator over 20 million euros.

Figure 5. Adding Balancing Weights. LVMS offers turbine experts online access to all data that are necessary for field balancing or turbogroups after a major overhaul.

This system is satisfactory and we now have a remote fleet wide monitoring solution that is more robust, quick, and easy to install and setup. We also have support from NI around the globe.


Author Information:
Kevin Dewaer
Engie Lab
Rodestraat 125
Linkebeek 1630

Bookmark and Share

Explore the NI Developer Community

Discover and collaborate on the latest example code and tutorials with a worldwide community of engineers and scientists.

‌Check‌ out‌ the‌ NI‌ Community

Who is National Instruments?

National Instruments provides a graphical system design platform for test, control, and embedded design applications that is transforming the way engineers and scientists design, prototype, and deploy systems.

‌Learn‌ more‌ about‌ NI