Wiseline predicts the future
Power grid maintenance entering the digital age
Wiseline tool enables power grid companies to understand their cost and risk drivers enabling them to optimise their maintenance program.
Every year there are thousands of both planned and unplanned interruptions on the electrical power grid in Norway. These disconnections result in undelivered energy that cost the society millions each year. The year 2016 saw a total of 25,777 incidents totaling a loss of 8,239 MWh undelivered energy (Statnett, 2017). Assuming an average interruption of 1.3 hours (Kjølle, 2011), the cost of undelivered energy is approximately 30.7 NOK per kWh. This equals 253.2 million NOK, or over 32 million USD (Ref. ).
Wiseline, with the support of among others Safetec, have worked to utilize the vast amount of inspection data into decision support. The result is a ground-breaking model empowering the power grid company to decrease both the number of power grid interruptions and saving inspections cost at the same time. This is achieved by two main steps;
1) Making the knowledge represented by the maintenance data available. Through an interactive dashboard where the power grid company gets a full overview of the risk related to each line.
2) Looking into the future, modeling a variety of maintenance programs to achieve an optimized solution combining reduced downtime and maintenance cost.
This project represents yet another project where Safetec combines risk management and digitalization competence, and the in-depth field experience of the client to create value.
(Click on the pictures for a close-up)
Picture 1: The screenshot presents a part of the risk visualization developed by Wiseline and Safetec. This is a PowerBI report representing each grid point with a dot. Given a green color, the technical state of the grid point is satisfactory, yellow is intermediate and red means that repair is needed. Note that all the boxes on the right side are filters. In this case the user has chosen the risk model “Technical state” for the year 2028, given maintenance strategy 6 (no maintenance). All the grey boxes are optional filters enabling the end user to choose a specific line, commercial zone, etc. The report is intuitively to use, within second the user can get an overview over the risk today (2018) and into the future given different maintenance strategies. NOTE: All data presented is fictional and does not represent the actual state of the grid points.
Picture 2: A part of the visualization PBI report is a table presenting the maintenance cost (the Cost column) and the state of the line. In this example, one can see that line 904 has an acceptable state of 98.53 in 2024 related to HMS given maintenance strategy 4 (see red filters). This has however cost 103 500 NOK to achieve. NOTE: All data presented is fictional and does not represent the actual state of the grid points.
Picture 3: The screenshot is the same report as in picture two. However, the end user has swapped to maintenance strategy 6 (no maintenance). Given no maintenance, the maintenance cost has dropped to zero, but the state of line 904 has also dropped down to the red zone with HMS state of 97.88. Hence the cost is reduced, but the probability of incidents has increased. NOTE: All data presented is fictional and does not represent the actual state of the grid points.
Picture 4: In earlier illustrations we have seen the cost per line given different maintenance strategies. It may be of interest to the end client not only to know the cost of a given strategy but also which failure modes result in highest maintenance cost, i.e. the cost drivers of the line in question. The illustration above shows that the combined cost for all years (all boxes in the filter “Year” is marked) given maintenance strategy 5 (repair when broken) is dominated by damage caused by woodpeckers and rot. NOTE: All data presented is fictional and does not represent the actual state of the grid points.
Picture 5: The illustration above presents an overall risk picture for all lines. In this case the end user has chosen the year 2026 and maintenance strategy 5 (repair when broken). In addition, the end user is interested in knowing the state of the most critical line financially (downtime cost the most) and has marked zone 4 in the filter. The illustration shows that all lines in 2026 seem to be in an okay technical state, however line 929, 959 and 966 are in the yellow zone. NOTE: All data presented is fictional and does not represent the actual state of the grid points.
 Edwin NJ, Mjølnerød H, and Gran BA; Data-Driven and Risk-Based Decision Support for Maintenance Planning on Electrical Power Grid Systems