Automatic Condition Monitoring in the Swedish Power Transmission System Based on Data Analysis
Doctoral thesis, 2024
Following a campaign in which all energy meters in the Swedish transmission grid were replaced, over 40 high-quality, instantaneous values are continuously collected from each energy meter every three seconds. In 2014, a research project was launched to investigate how this data could be used. This thesis covers selected results from a journey of over ten years, using real-world data to improve the quality and efficiency
of the Swedish transmission grid’s maintenance (asset management) and operation. It refers to this process as automatic condition monitoring (ACM). Most of the results from the research are now in daily use by Svenska kraftnät, the state-owned transmission system operator in Sweden (Svk), in its operations. Important results in this thesis includes:
• Development of a general methodology for ACM on how to use substation measurements. This methodology has been used to successfully develop algorithms that can detect specific equipment problems, as well as algorithms that detect previously unknown issues.
• Specific algorithms for automatically detecting individual punctures in capacitor voltage transformers (CVTs). These algorithms were verified over several years of real-world operation across the entire Swedish transmission system. The work led to lower personnel and operational risks, reduced maintenance costs and improved measurement accuracy in the revenue metering process. Over 74 individual punctures were detected using these algorithms. As of October 2024, Svk is monitoring 59 active punctures every day to ensure that the CVTs will not degrade further and explode.
• An in situ calibration algorithm for energy meters. This algorithm uses measurements from energy meters and can simultaneously detect non-linear measurement errors across all energy meters in a substation. It now forms part of an accredited verification process which is regularly used in the transmission system.
• A new SF6 gas leakage detection algorithm has been developed, which offers unprecedented performance. This algorithm can detect small leaks within 15-25 days, unlike other reported algorithms that can take a year (or more) to accomplish the same task.
• Following a thorough investigation regarding the challenges and opportunities of implementing the process bus and digital metering into the substation design, a new solution has been introduced that offers an efficient digital substation metering solution. The proposal is based on virtualising all accurate metering instruments and hosting them all in a single physical unit, rather than installing several tens of physical instruments in the control building.
of the Swedish transmission grid’s maintenance (asset management) and operation. It refers to this process as automatic condition monitoring (ACM). Most of the results from the research are now in daily use by Svenska kraftnät, the state-owned transmission system operator in Sweden (Svk), in its operations. Important results in this thesis includes:
• Development of a general methodology for ACM on how to use substation measurements. This methodology has been used to successfully develop algorithms that can detect specific equipment problems, as well as algorithms that detect previously unknown issues.
• Specific algorithms for automatically detecting individual punctures in capacitor voltage transformers (CVTs). These algorithms were verified over several years of real-world operation across the entire Swedish transmission system. The work led to lower personnel and operational risks, reduced maintenance costs and improved measurement accuracy in the revenue metering process. Over 74 individual punctures were detected using these algorithms. As of October 2024, Svk is monitoring 59 active punctures every day to ensure that the CVTs will not degrade further and explode.
• An in situ calibration algorithm for energy meters. This algorithm uses measurements from energy meters and can simultaneously detect non-linear measurement errors across all energy meters in a substation. It now forms part of an accredited verification process which is regularly used in the transmission system.
• A new SF6 gas leakage detection algorithm has been developed, which offers unprecedented performance. This algorithm can detect small leaks within 15-25 days, unlike other reported algorithms that can take a year (or more) to accomplish the same task.
• Following a thorough investigation regarding the challenges and opportunities of implementing the process bus and digital metering into the substation design, a new solution has been introduced that offers an efficient digital substation metering solution. The proposal is based on virtualising all accurate metering instruments and hosting them all in a single physical unit, rather than installing several tens of physical instruments in the control building.
energy measuring systems
measurement errors
algorithms
supervision algorithms
Kalman filter
CVT
leak detection
digital substation
capacitor voltage transformer
SF6
IEC 61850
capacitor puncture
monitoring
health
circuit breaker
unbalance
process bus
digital metering
asset management
CCVT
Lecture hall EF, Hörsalsvägen 11, Gothenburg.
Opponent: Professor Lars Nordström, KTH, Sweden.
Author
Anders Lindskog
Chalmers, Electrical Engineering, Electric Power Engineering
This thesis presents new ways to detect degrading equipment in power transmission systems by using automatic condition monitoring.
Svenska kraftnät is responsible for maintaining, developing and operating the Swedish power transmission system. For a safe operation, it is important that all equipment in the system works according to its specifications. Thus, the condition of the equipment is regularly monitored and maintained before it may fail and threaten the operation of the grid.
Traditionally the condition monitoring of the equipment is manual and calendar based. However, degradation can occur rapidly and the equipment can potentially fail between inspection visits, causing operational risks and substantial costs. For these reasons, it is favourable to use automatic condition monitoring to continuously monitor the equipment and alert maintenance personnel when degradation occurs.
The presented work shows how substation measurements, with a focus on electrical measurements, can be used for condition monitoring. The thesis gives examples of monitoring methods that have been developed and are in use in the Swedish power transmission system. It also provides information on useful prerequisites for performing automatic condition monitoring.
The target groups for the thesis are transmission system operators and other industrial and academic stakeholders. Hopefully, it will provide useful information for anyone wanting to follow a similar path to that of Svenska kraftnät.
Svenska kraftnät is responsible for maintaining, developing and operating the Swedish power transmission system. For a safe operation, it is important that all equipment in the system works according to its specifications. Thus, the condition of the equipment is regularly monitored and maintained before it may fail and threaten the operation of the grid.
Traditionally the condition monitoring of the equipment is manual and calendar based. However, degradation can occur rapidly and the equipment can potentially fail between inspection visits, causing operational risks and substantial costs. For these reasons, it is favourable to use automatic condition monitoring to continuously monitor the equipment and alert maintenance personnel when degradation occurs.
The presented work shows how substation measurements, with a focus on electrical measurements, can be used for condition monitoring. The thesis gives examples of monitoring methods that have been developed and are in use in the Swedish power transmission system. It also provides information on useful prerequisites for performing automatic condition monitoring.
The target groups for the thesis are transmission system operators and other industrial and academic stakeholders. Hopefully, it will provide useful information for anyone wanting to follow a similar path to that of Svenska kraftnät.
Areas of Advance
Energy
Subject Categories
Electrical Engineering, Electronic Engineering, Information Engineering
ISBN
978-91-8103-142-3
Doktorsavhandlingar vid Chalmers tekniska högskola. Ny serie: 5600
Publisher
Chalmers
Lecture hall EF, Hörsalsvägen 11, Gothenburg.
Opponent: Professor Lars Nordström, KTH, Sweden.