CoDiS Machine Condition Monitoring System Installed on 185 MW Pumped-Storage Hydropower Plant

  Print Print

"We installed the state of the art CoDiS system, a solution based on NI technology, for machine condition monitoring, on a large (185 MW) variable speed pumped-storage hydrogenerator."

- Ozren Oreskovic, Veski d.o.o.

The Challenge:
Building a customized online machine condition monitoring (MCM) solution to track relevant and significant changes in machine behavior for 185 MW unit in a pumped-storage hydropower plant.

The Solution:
Installing an online monitoring, protection, and database storage system to integrate the measurements of many different quantities covered by different hardware and software modules, designed to monitor: rotor vibro-dynamics (relative shaft and absolute bearing vibrations on three lead bearings), turbine cover vibrations, generator air gap and rotor pole magnetic field, axial displacement of rotor on two positions, thrust bearing clearance (four sensors per bearing in radial direction), thrust bearing temperatures, electrical quantity monitoring (power quality vector monitoring, electrical transient recording), process quantities received from SCADA (54 values overall including temperature of stator, bearings, and operating condition status), partial discharge, cavitation, hydraulic quantities, 32 values (pressures, flows, relay conditions) of transient recordings.

Author(s):
Ozren Oreskovic - Veski d.o.o.
Franjo Tonkovic - Veski d.o.o.
Ozren Husnjak - Veski d.o.o.

An online monitoring system installed in a power generating plant can continuously monitor operating conditions, which is a necessary step toward predicting mechanical behavior. The real benefit is using malfunction prediction to detect and identify irregularities in system behavior, which improves repair planning, overhaul planning, and scheduling for additional damage occurrence prevention. Up-to-date condition monitoring systems include vibro-dynamic and generator air-gap monitoring, and monitoring of electrical and process values (temperature, pressure, flow, and more).

A pumped-storage hydropower plant is the biggest hydropower plant in Slovenia and very important for energy stability improvements. During periods with low electric energy prices (nights and weekends) the plant pumps the water into the accumulation reservoir. During periods of high electric energy prices (the peak periods on weekdays) accumulated water can be used to generate and sell electric energy.

Consequently, the machine has four different operating modes (generator steady, generator condenser, pump, and pump condenser) and usually has 3-4 daily run-ups and coast-downs, which makes it challenging to continuously track data. Continuous machine health tracking is very important as the plant has no on-site crew and is completely operated from a remote dispatch center. A machine condition monitoring (MCM) system is available on the plant and dispatch network, and CoDiS client software brings all the data to remote user PC locations.

The system features a protection module for vibration and mechanical values (CoDiS RT), a diagnostic module with SQL database (CoDiS DM), electrical monitoring with power quality (CoDiS TR), and hydraulic transients monitoring (CoDiS-CV as a part of DM module). The user gains complete insight into machine operating condition, the quality of generated power, and all significant data constantly stored to database for history overview, trend display, and offline analysis.

VESKI, a Croatian-based company that specializes in MCM including vibration analysis and diagnostics, air gap and magnetic flux monitoring, hydraulic quantities monitoring, relies on LabVIEW software and CompactRIO hardware as the core of its CoDiS monitoring platform. We based the CoDiS MCM system on CompactRIO processing hardware and the LabVIEW platform as a base of analysis, diagnostics, and visualization. The system is installed in more than 40 power plants worldwide. The system in this plant includes four self-standing standard main monitoring modules:

  • CoDiS RT—Online real-time/FPGA protection
  • CoDiS DM—Continuous diagnostic monitoring with database and remote client access
  • CoDiS TR—Online electrical transient and power quality monitoring
  • CoDiS CV—Online hydraulic quantities monitoring module

We based the CoDiS RT and CoDiS TR modules on CompactRIO rugged technology because we needed them to run constantly in power plants or similar industrial environments. In the Avče pumped-storage plant, we installed the CoDiS RT and CoDiS TR in the same configuration with a cRIO-9024 RT controller and a cRIO-9113 FPGA chassis. We used the NI 9205 and NI 9234 for data acquisition and the NI 9481 for relay protection of the machine. The CoDiS DM module communicates with the CoDiS RT and CoDiS TR modules over TCP/IP, transfers the online data to the user, and stores the data in a SQL database for offline analysis and history tracking.

Real-Time Protection CoDiS RT

CoDiS RT delivers high-priority FPGA protection, real-time protection, and 45 input channels with a standard (configurable) sampling rate of 2,048 Hz per channel. The system can recognize the operating mode (based on RPM and active and reactive power tracking) and adjusts the protection functionality (thresholds and delay times) to the specific regime. This is very important because the vibration and hydraulic response is different in generating and pumping regime, so the values for alarm and trip must be adjusted automatically.

The system continuously acquires all signals, constantly calculating significant values from the waveforms of the input signals. Calculated values (forming a condition vector) are compared with the protection settings for that specific regime. In cases of irregularity, the system triggers the relays and shuts the machine down. Synchronously, all the data are transferred to server PC for further analysis and data storage. The user can set all the values by means of the administration software with administrator permissions.

Electrical Transient Recorder CoDiS TR

The CoDiS TR module measures 22 electrical quantities with a 5,000 Hz sampling rate per channel. The system continuously acquires all signals and compares the trigger conditions for fast electrical transients. In cases of transient occurrence, the system streams the data to hard drive in a duration of 10 s with 1 s of pre-triggering. The power quality analysis procedures are applied on waveform signals on the CompactRIO. Calculated data is sent synchronously to the server where the power quality measurements condition vector is stored to one unique database for all measured quantities.

Diagnostic System CoDiS DM

The diagnostic system runs on a server and communicates with distributed real-time processors on one side and with users on the other. CoDiS DM receives data from real-time systems every second and recognizes the adequate storage type adjusted to the specific machine operating mode.

  1. System recognizes the operating mode (generator or pump)
  2. System examines the data to determine the type of analysis (steady state or transient)
  3. In case of steady-state mode, the system averages the data and stores it to a trends database in a time-averaged period determined by the user (for hydro machine, the average is every minute or two)
  4. In case of transient mode, (run up/coast down) the system records data with full resolution
  5. In case of event occurrence, the system automatically stores the waveforms, performs additional analysis of the signals, and stores this data in the event database (either alarm-automatically or user-on-demand created)

At the same time, the DM module distributes the real-time data values to clients if connected. The server also communicates with SCADA, exchanging the data every second so the database also contains the process quantities such as water head, stator temperatures, pressures, and more. The database for electrical and mechanical quantities is unique and the system is synchronized on a plant internal GPS clock, so data comparison with SCADA and DCS can be done with the same time stamp

CoDiS DM—Standard Data Analysis Tools

For mechanical measurements:

  • Bode plot, Nyquist plot
  • 2D and 3D orbit analysis
  • 2D and 3D shaft centerline
  • Run-out compensation
  • Air gap polar plot and stator geometry
  • Magnetic field pole profile and polar display
  • FFT spectrum, CPB spectrum
  • Waterfall and cascade spectrum

For electrical measurements:

  • Transient detection with vibration signals
  • Power quality
  • Generator load angle
  • Torsion magnetic field stiffness

CoDiS DM—Expert data analysis tools:

  • Bearing and structure stiffness identification
  • Critical speed identification
  • Statistical analysis of machine operation
  • Torsion vibration identification
  • Automatic fault detection and prediction

Non-Standard Monitoring Modules CoDiS CV

MCM Avče includes some additional modules:

  • Hydraulic value transient module CoDiS CV is installed in addition to CoDiS DM (32 values, two locations with sampling rate of 1,024S/s, 1 min pre-trigger and 5 min post-trigger automatic and unlimited on-demand recording)
  • Lower guide bearing bracket deformation and temperature monitoring
  • Upper guide bearing bracket deformation and temperature monitoring

Remote Client Software

Client software is part of the diagnostic system and runs on user remote PCs or is accessible as a web application. The client interface includes a hierarchy plant view with the most important values on screen, including the alarm status. Machine details are given in online detailed overview panels with all real-time measurements displayed (this module is most commonly used by operators in dispatch centers). It is considered very useful when the machine is started remotely. The plant engineers use the database and analysis software for maintenance planning and irregular behavior detection.

Automated Reporting and Fault Detection

To support the customer in making the right maintenance decision, VESKI offers remote consulting and support and can install an automated reporting and fault detection module. We developed these tools with a knowledge base for hydrogenerators that was accumulated by partnering with maintenance engineers from many plants throughout the world. The system can detect the fault and predict the failure in the early stage of the fault appearance. Results are displayed as a message to the user with a detailed fault description (fault location, type, and probability). The system can generate automated reports to email to the user on a daily basis. In HPP Avče, the automatic report generation module is activated so the user receives the most important data extracted from the database on a daily basis. The reporting can include both fault analysis and statistical analysis of data from a time period defined by the user.

Author Information:
Ozren Oreskovic
Veski d.o.o.
Oreskoviceva 8j
Croatia
Tel: +38513667133
info@veski.hr

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