Learnings

Outcomes

The main outcome of this project is that a new concept for extending state-of-the-art condition monitoring techniques to submarine cables of any length has been developed. This solution combines pre-existing technologies from the fields of distributed optical fibre sensing and fibre optic telecommunications to unlock a new level of performance; extending measurement range from circa 50 km to unlimited. 

Over the course of the project the concept has been developed (in line with the first success criterion) and is now fully formulated, meaning the concept has been advanced from TRL1 to TRL2. Our verification and validation efforts (in line with the first and second success criteria) represent experimental proof of concept and preliminary laboratory verification indicating TRL3 to TRL4. The tunnel-based field testing could constitute validation in a relevant environment, TRL5. While specifications for future versions of the proposed solution have been suggested, the third success criterion refers to BAU adoption (TRL9 or beyond) which will require considerably more investment to reach. 

The TRLs stated above (i.e. final level TRL4 to TRL5) apply to our laboratory apparatus which has been assembled from commercial-off-the-shelf (COTS) components (although highly specialised ones in some cases). The TRLs should therefore be caveated with the fact that it is unlikely that devices assembled from COTS components would be suitable for BAU adoption. It is therefore expected that BAU adoption would require complete re-engineering with custom electronics to reduce power consumption, heat and to maximise reliability and projected service life. In that context it may be more realistic to consider this project as an experimental proof of concept, with a final technology readiness level of TRL3.

Lessons Learnt

The experimental deliverables produced during this work period were D5, D7, and D8; the key learnings from each will be briefly summarised in the following paragraphs. 

Deliverable D5:
D5 is a technical report on initial experimental results from the laboratory test-bed. At the most fundamental level, obtaining these initial results demonstrates that the components selected and the in-house developed DOFS system are operational. More detail for some of the key findings is set out below:

• Section 2: Preparing the apparatus for use outside of an interlocked laser laboratory (e.g. for the tunnel-based testing) added considerable complexity to the design. I.e. A laser interlock system and complex electrical feedthroughs had to be added. Unfortunately, the enclosure also compromises heat dissipation.
• Section 3: The remote-control systems were outlined in the report and tested experimentally both with a LAN based connection (for the submarine sensor pods) and with mobile 5G networking (for the tunnel testing).
• Section 3: Secure remote control of the Arduino Opta (which has limited built-in security) was achieved by configuring the 5G router (which manages the remote apparatus’s internal network) as a VPN bridge.
• Section 4: The capability of the FPGA-based data acquisition card (ADQ14) for high-speed data averaging was tested and proved to be invaluable for improving the signal-to-noise ratio of the DOFS signal.
• Section 4: Some unexpected difficulties arose relating to laser triggered acquisition. Additional hardware was applied to improve grounding and isolation of the laser and ADQ14 but this did not completely resolve the issue. Current workarounds are to reduce the laser repetition rate (increasing the time needed for each measurement) and to apply additional data processing steps (such as “background subtraction”) 
• Section 4: The upper limit for laser pump power (due to the onset of stimulated Raman scattering) was empirically determined to be 27 mW (for 2 ns pulses at 40 kHz with a 1 km sensing fibre).
• Section 4: Initial observations of the temperature dependence of the Anti-Stokes Raman signal were made.

Deliverable D7:

The main focus of D7 is on the design of the power supply systems associated with the CALICS device.

• Section 2: The physical configuration of the cables and sensor pods that make up the CALICS and power cable systems have impacts on power delivery, heat dissipation and electromagnetically induced current flows.
• Section 2: The evolution of the CALICS concept through different configurations was set out. In the currently preferred implementation, CALICS uses a separate cable, running parallel to the power cable, which provides power and data connection to modular sensor pods which can be tailored to the specific application. Each pod provides fibre optic sensing (in one or more modalities) over a segment of the power cable and could also incorporate other local sensing systems.
• Section 3: Details the remote powering scheme utilised by submarine telecommunications cables, and includes calculations that show why low current, high voltage DC is favourable for long interconnects.
• Section 3: Despite their relative inefficiency, Zener diodes are favoured as a high reliability method of extracting power from the constant HVDC flowing along the single conductor cable.
• Section 4: The sensor pod power consumption was estimated as 170 W (upper limit) and 5 W (lower limit) based on the chosen components and the efficiency of the various DC-to-DC converters. These two extremes must both fall within the combined operating window of the Zener diodes used. A system based on 8 series connected 25 W Zener diodes (giving a nominal voltage drop of 200 V) was found to be optimal.
• Section 4: Power consumption of the selected components was experimentally verified and in normal operation was 50 % lower than the estimated upper limit, leaving a comfortable safety margin.
• Section 5: Estimates were made of the voltages induced on a 50 km long section of CALICS cable running parallel to power cables with different magnitudes of AC ripple. At a separation of 10 m, induced voltages are only expected to be problematic for AC power cables, or HVDC systems with LCC converters and >1 % ripple.
  Section 6: The cybersecurity risks associated with CALICS vary significantly depending on how it is configured.If the CALICS cable data network is isolated, then due to physical inaccessibility its most vulnerable parts would be the computer terminals at the cable landing stations. The risks progressively increase as more features are enabled (e.g. automatically uploading data to a web connected data server, allowing sensor pod configuration via the web interface, deeply integrating the CALICS data into the power cable SCADA system).

Deliverable D8: 

The activities defined under Deliverable D8 were undertaken to the extent necessary to assess the feasibility and value of conducting a full tunnel-based field trial of the CALICS apparatus at the Elstree to St. John’s Wood cable tunnel.

Preparatory considerations were completed, including initial planning for training, risk assessment, access requirements, and logistical arrangements. A survey-based review was also carried out to identify potential installation locations, assess viable cable routing options, and consider practical installation approaches and attachment methods. These activities enabled a comprehensive understanding of the deployment requirements and constraints within the tunnel environment.

In parallel, the expected performance outcomes of the CALICS apparatus had already been robustly established through prior laboratory testing and analysis, including assessment against relevant load data. As a result, the incremental technical insight expected from a full-scale field trial was determined to be limited.

Following discussions during Workshop D9, it was concluded that proceeding with the tunnel trial would require significant additional cost, effort, and access coordination, while delivering minimal additional value beyond what had already been demonstrated. In consideration of these factors, and to ensure responsible use of taxpayer funding, the decision was taken not to proceed with the physical field trial.

Consequently, Deliverable D8 has fulfilled its objective by confirming the feasibility of deployment, validating the expected outcomes through existing evidence, and demonstrating that a full-scale trial would not provide proportionate additional benefit.

Dissemination

EPE RAD 25: An overview of the CALICS project was presented to members of the University of Southampton, Electrical Power Engineering group on the 3rd of April 2025.

 
Deliverable D9:

A workshop was held at the University of Southampton on 18th of February 2026 with several representatives from NG attending in-person and online. Presentations explained the CALICS concept, its objectives and the developments and key findings that have been made throughout the project. Recommendations and specifications for implementing CALICS were outlined. For example, the system should make use of commercially available telecommunications power feeding equipment (PFE) as this would allow the inclusion of standard fibre optic repeater units, if necessary, on the longest cable systems. There is potential for significant improvements in power consumption and heat dissipation and physical size if custom electronics are used rather than commercial off-the-shelf (COTS) parts. Investigating the inclusion of distributed acoustic sensing (DAS) within the sensor pods is highly recommended, as this has great potential for security monitoring of submarine infrastructure, to compliment the DTS condition monitoring data. 

Wider dissemination beyond project partners has not yet been pursued as there could be potential for commercialisation of the CALICS concept but strategies and policies for this have not yet been agreed.

Discussions under NDA have been held between UoS and SubConnect (part of the Global Marine Group) and an agreement has been entered into for supply of a reclaimed repeater pressure vessel, lengths of submarine fibre optic cable and services and training in opening/closing and installing our CALICS hardware within.