In the last part we looked at how we can use Mermaid outputs to perform a bespoke analysis and discussed the process of writing a script to extract additional information. This example is reasonably simple, but the timestep by timestep nature of the results allows substantial detail to be considered.
In this post we’ll take a look at the results we generated by running this script on our cases, considering only simulations starting in June. We’ll also consider the process we have undertaken and the impact that this type of analysis can have on the success of a project.
Presented below are the results for the base case, where the offset between the trencher being hired and the cable lay vessel being hired is 32 days. Firstly, an explanation of the chart, which was produced in a spreadsheet, rather than directly in Mermaid, and is repeated through this post.
All simulations are displayed relative to the cable lay vessel being taken on hire (this is 0 on the x-axis). The blue dot represents the point at which the cable lay vessel went off hire and the green bar represents the period of time for which the trencher was on hire (from 32 days after the cable lay to the completion of its operations). Each row represents one simulation.
Generally a preferable operation will have:
- The Cable Lay off hire point as far to the left as possible, reducing the period of time this vessel is on hire for.
- The Trencher hire period bar as short as possible, reducing the period of time this vessel is on hire for.
Also, we may want the latter of the vessel off hires as far to the left as possible, meaning cable installation is completed in as short a period of time as possible.
When we consider taking the Trencher 3 days earlier we see that the Cable Lay vessel is unaffected (we wouldn’t expect it to be). We also see that the Trencher finished at the same point as in the base case and that all we have achieved is extending the hire period. This is as we expected when we considered the results in Part 3, but it is good that we have confirmed this.
In fact, taking the vessel any amount earlier than our base case shows the same trends (see further below for taking it 28 days earlier) – that the operation is no more successful, but that our hire times have increased.
If we take the Trencher 3 days later we see that:
- The Cable Lay vessel is again unaffected (again this is not surprising as it is not dependent on the trenching vessel).
- In most cases we see that the Trencher finishes at the same time as in the base case, although in some cases the Trencher now finishes work after the Cable Lay vessel. This extends the time to cable installation completion, but reduces the hire time of the trencher by reducing waiting at the start.
Indeed, we see this pattern again in the case where the Trencher is taken on hire 7 days later. We have once again extended the duration of the project we have reduced the waiting incurred by the trencher.
Finally, as we suspected earlier, as we go even later with the trencher we totally decouple the relationship between the two. Selecting this type of scheduling will be highly dependent on:
- The acceptability of leaving laid cables un-trenched for an extended period of time.
- The acceptability of increasing the overall project length.
We’ve looked at the cases individually, now we should consider which schedule we prefer. The chart below shows the maximum, minimum and median duration, in days, for the project (blue line) and the trencher hire period (green). We want to minimise both of these as much as possible but will likely be faced with a trade-off between a slight increase in project length compared with a reduction in hire time. This decision is likely to be project specific. We could analyse this operation even further to support this decision making, but we do have here data which provide the required tools and insight to inform this decision. Considering only the data presented here, taking the trenching vessel 7 to 10 days later (39 to 42 days after the cable vessel) appears preferable.
Summary and conclusions
We’re aware that weather risk is a very complicated issue, but through detailed and careful modelling we can get a good grip on where the risk lies, and where it doesn’t. In this example we’ve only looked at one variable, the time at which we want to take the vessel on hire, but there are further issues we could investigate, for example:
- What about the other components of the operations:
- The ports used;
- The vessels used;
- The trencher capability;
- The cable lay speed;
- The trenching speed;
- Optimisation of pull-in;
- And many more.
We may also want to consider operations and maintenance, and decommissioning.
In this example we determined that our base case was a fairly good scenario, but we still managed to optimise it at a saving, in day rate terms, of many tens, or hundreds, of thousands of pounds. We can also go beyond where we have got to here and start to think about when energy needs to be delivered to the grid and the value of delivering energy earlier, even at greater expense. More detailed analysis might lead to us generating revenue significantly earlier if we understand where to direct our efforts. For example, the delay to cable laying to reduce the vessel expenditure may not be critical and therefore acceptable, whilst another operational element may require further consideration.
Throughout this series we saw a number of ways to model operations in Mermaid, and a number of way to consider the results of our analyses, including bespoke post processing. This is a hypothetical example but the steps involved can easily be applied to your analyses. As always, please do get in touch with any comments or questions.