Dynamic Cable Model

Liander N.V.

The dynamic cable model is a model that can accurately calculate cable temperatures of medium- and high-voltage cables in changing conditions. Here, environmental situations such as ground temperature, soil moisture and load profile can be dynamic over time, allowing realistic situations for cables to be calculated.

Last change on 17th of October 2024, at 14:03 (CET) | Publication Standard 1.0

• Publication category: Impactful algorithms
• Impact assessment: Field not filled in.
• Status: In use

General information

Theme

Space and Infrastructure

Begin date

2022-01

Contact information

[e-mail]

Responsible use

Goal and impact

Anno 2024, there is a lot of congestion on the electricity grid. More and more customers want to electrify, which puts enormous pressure on electricity grids that are not designed for this. National and regional grid operators are therefore looking for the space we have on the electricity grid what we already maintain. With this model, the dynamic cable load capacity (maximum load capacity) can be determined. By taking into account the anonymised and aggregated load profiles of our customers and the environmental variables of our power grid, we can ensure that we can connect more customers during the energy transition. The dynamic cable load capacity can be much higher than the load capacity set by the grid operator's policy or by the cable manufacturer, because it is not based on generic conservative values, but on the actual situation. This model has been developed to allow the grid operator to deviate from their conservative standards so that more customers can be connected to the grid.

Considerations

Use of this product supports the organisation within its digitalisation strategy and allows us to remotely determine whether we can connect more customers to the grid based on physical models. This allows us to minimise the deployment of expensive sensor solutions and enable the energy transition within accepted and substantiated financial and safety risks.

Human intervention

Humans always make the choice about what is done with the outcome of the model. Specialists with knowledge of the grid use this model to input data and ask for an outcome, with this they then do/do not work further depending on the outcome. There is always a human making the decisions based on the model's calculations.

Risk management

Risk management for the DKM algorithm consists of several components: the safe limits it computes with; human intervention; logging of inputs and outputs; and audits are performed on the algorithm.

Safe limits

Liander's policy defines limits within which the algorithm makes calculations. The algorithm also gives more general advice when the little information is provided and more specific results when more information is provided, the results always remain within the set safe margins. This ensures that unsafe situations cannot arise due to calculation errors.

Human intervention

In addition, it is not automated decision-making, a specialist enters data, with this the algorithm makes a calculation. Then the specialist looks at the outcome and if/how to proceed with it. The specialist has knowledge of the grid and can therefore determine not to use an outcome. So this model cannot autonomously cause errors in the grid, there is always a person with specialist knowledge in between who uses the model's output.

Logging

All data entered into the algorithm as input is stored, this way we can, if necessary, reason back to what a particular outcome is based on.

Audits

An internal audit was carried out on the algorithm in order to identify possible risks of deploying it. The conclusion of this is that the model is soundly designed, partly because of the clear documentation of components that are or are not included and what their impact is. They also mention that the validation of the calculations is well set up, and that the validation of the reliability of the inputs is thorough and well documented.

Finally, the underlying systems used are patched in a timely manner to prevent unauthorised access.

Operations

Data

The dynamic cable model uses various data sources to make calculations. The data is retrieved at the moment a user wants to do a calculation and the model then uses the most recently available data. The dynamic cable model uses the following data sources in its calculations:

1. ERA5 Weather Data - an open dataset storing real-time weather information.

2. BOFEK2012 subsurface model - an open dataset containing soil physical characteristics of soils throughout the Netherlands.

3. LHM groundwater model - the national hydrological model is an open dataset and provides information on various characteristics of groundwater in the Netherlands.

4. Liander Asset Data - dataset from Liander that includes assets placed in the Netherlands. Used to see what type of cable is in which place, for example.

Links to data sources

• LHM: https://nhi.nu/modellen/lhm/#:~:text=Het%20Landelijk%20Hydrologisch%20Model%20is,in%20verschillende%20overwegend%20landelijke%20studies..
• ERA5 weerdata: https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels?tab=overview
• BOFEK2012: https://research.wur.nl/en/publications/bofek2012-de-nieuwe-bodemfysische-schematisatie-van-nederland

Technical design

The input data mentioned in the Data section are converted by logic and physical calculations into the maximum cable load capacity based on the temperature of the cable. Using information about a power cable and information about external factors affecting the temperature evolution of power cables, the model calculates the temperature a cable reaches when a given amount of current flows through the cable. Relevant information about the cable includes, for example, its length and material. Examples of external factor information are weather information and information about the soil composition where the cable is located. This external information determines how fast a cable heats up given the conditions it is in, which allows us to run more current through the cable than the prescribed maximum. A simple example is: In winter, the ground is colder, so cables heat up less quickly and more current can flow through them.