DVL aiding – in the harshest environments

In pipeline inspection operations, for example, it’s preferable to get the vehicle as close to the seabed – and the pipeline it’s inspecting – as possible. But, this can cause the DVL’s bottom lock to be compromised, due to being too close to the seabed.

The challenge

For a number of operations, getting survey and positioning sensors as close to the seafloor as possible is important. However, this can lead to special requirements for sensor and vehicle design and an increase in operational complexity.  Mounting the Doppler Velocity Log (DVL) with an unobstructed downward view, to achieve bottom lock, can become particularly difficult.

When operating ploughs or trenchers, most mounting options – eg. pointing the DVL directly towards the seabed, as is commonly done – compromises the DVL’s accuracy, due to the highly turbid water created by the vehicle it’s mounted on and the DVL beams being reflected or blocked by the trencher or plough itself.

One option, in these applications, is to separate the DVL from the inertial navigation system (INS). This allows more flexibility in terms of where the DVL is mounted. However, this can also adversely impact navigational accuracy, as it introduces lever arm and mounting angle errors between the INS and the DVL. Furthermore, performing DVL to INS calibrations are often impossible or very impractical on vehicles such as ploughs and trenchers, because vehicle manoeuvrability is not suitable for calibration purposes.

A solution for DVL mounting that improves flexibility on integration without compromising performance is required.

The solution

When mounting a DVL, one of the limitations is that optimal performance often requires vertical mounting. Our SPRINT-Nav – a compact INS and DVL in one – removes this limitation as it can be mounted at an angle. This avoids turbidity and the vehicle’s own infrastructure, which would otherwise interrupt its DVL beams.

The SPRINT-Nav can do this because the DVL it contains is tightly integrated at beam level. Each individual DVL beam is fed directly into the INS solution, instead of feeding in a 3D velocity solution that the DVL resolves by itself. This low-level integration means that SPRINT-Nav does not rely on the DVL alone for resolving a 3D velocity solution – it is handled more optimally by the INS. The SPRINT-Nav is then constrained and, critically, can be mounted at an angle, without compromising accuracy.

The idea of mounting the SPRINT-Nav at an angle was initially raised during a meeting with offshore survey solution provider UTEC. The concept instantly sparked great interest, as it had not previously been possible with existing INS solutions.

The next step was to test and evaluate how the alternative SPRINT-Nav mounting arrangement would perform for this specific application. The operation chosen for the evaluation was a cable lay project in the East Anglia ONE offshore wind farm, in the UK’s southern North Sea. There, UTEC was working from DeepOcean’s Maersk Connector cable lay vessel, providing positioning of DeepOcean’s plough during simultaneous lay and trench of a cable. UTEC’s objective was to get an accurate INS/DVL based position for the plough in order to replace the USBL positioning which had been degraded by acoustic noise from the vessel thrusters.

The project included both shallow and deeper water operations, with some very challenging conditions – highly turbid water, changing altitudes from the varying plough depth and sand waves and a very high level of vibration. Additional considerations included varying speeds as the plough made its way along the seabed and intermittent, noisy USBL updates from long layback tracking and vessel thrust.

All-in-all, it was a very tough application, in a very demanding environment. Both the inertial solution and the acoustics were thoroughly put through their paces. All the same, we were confident that SPRINT-Nav would perform well, despite the challenges it faced.

The ring laser gyro (RLG) technology used within SPRINT-Nav is almost immune to vibration; this is not the case for other competing gyro technologies. The tight beam-level DVL aiding increases robustness, so even if the DVL loses bottom track periodically or loses individual beams high levels of accuracy and reliability are maintained. What’s more, the pre-calibrated SPRINT-Nav does not rely on any additional calibrations or fine-tuning to ensure performance and the robustness of the SPRINT-Nav solution makes it less dependent on regular USBL updates.

This mounting arrangement had two benefits. Firstly, the SPRINT-Nav would always have four DVL beam returns as opposed to mounting it vertically, which would cause at least one of the beams to bounce off the plough’s infrastructure. Secondly, three out of four DVL beams would be directed away from the most turbid water surrounding the plough. This improves the signal-to-noise ratio and provides a more robust INS solution.

A bespoke mount was manufactured for this project. This enabled us to adjust pitch and yaw angles when needed during testing. Furthermore, post-processing in our Janus software was proposed to deliver the most accurate as-laid cable data.

The cable lay project was divided in three parts: integration and mobilisation; shallow water shake-down; and ploughing. A Sonardyne engineer was present for the initial shallow water shake-down and after a review of the integration, the system was deployed.

Throughout, SPRINT software was used to monitor the raw beam velocity data from the SPRINT-Nav’s integrated DVL. It became instantly obvious that all beams were achieving a good lock, even with the 30° pitch angle. It was also seen that the system continued to observe good data throughout the entire operation.

The results

Throughout both the shallow water and deeper water cable lay operations, the SPRINT-Nav performed well and no performance degradation from the 30° pitch angle mounting was experienced. On-board team members also found the topside software (SPRINT) intuitive and very useful for troubleshooting and tweaking the system.

As the environment was very new to Sonardyne (low speed, jerking motion, high turbidity), we also collected a lot of very useful data. This included DVL and INS data, which has helped us to understand how we can further improve the set up for these and similarly challenging applications.

We thank UTEC and DeepOcean for supporting this trial and look forward to providing this solution on future projects.