Last time we finished the setup of the analysis by preparing the flow diagram. We considered a number of options for modelling these tasks, ultimately selecting a clean, simple and easy to maintain approach. We also performed a number of simulations.
In this post we will take a look at the results we have generated and will consider the impact these can have on our decision making. Finally, we’ll make some decisions regarding the next steps we should take.
Box Whisker Charts
In this post we are going to analyse our results using Mermaid’s “Duration Box Whisker Charts”, here’s a quick overview of the data represented by different elements of these.
In the image above we can see a number of data items. Firstly, the horizontal red line at the bottom of the chart represents the “Unweathered” case. This scenario is the operation performed without the impact of any metocean conditions, basically, this represents perfection for this plan.
The red box represents the interquartile range (P25 at the lower edge to P75 at the upper edge), the horizontal line across this box represents the median duration. The end of the lower whisker (vertical line) is the minimum duration and the end of the upper whisker is the maximum duration. All results for this example are for simulations which start in January.
Presented below is the box whisker plot for the base case scenario (where we take the trenching vessel on hire 32 days after the cable lay vessel). We can note a number of points from this:
- The unweathered duration is 1114 hours (46.4 days). We will not be able to perform this operation any faster than this.
- In some spring/summer months we are able to perform the operations without the impact of weather, i.e. conditions are good and we are unaffected.
- Our winter durations are highly variable and give us little confidence in our ability to perform the work in a given time.
- The spread of durations in April, May and June is low; we have a more dependable operation (i.e. we can be pretty sure we’ll get the operation completed in around 60 days).
Since the operation only takes around 60 days it is sensible to perform it in the better weather of the northern hemisphere summer (i.e. start in late spring/early summer). This isn’t always the case, as can be seen in this interesting series of posts.
Changing vessel on hire time
We also performed simulations for a number of different on hire offsets, these are recapped in the table below.
Firstly let’s look at taking the vessel on hire much earlier (-14 days and -28 days). Shown in the box whisker diagram below are the results for these two cases, run from March to August, and the base case. Here we can see that:
- There is no change to the unweathered case. This means that taking the vessel on hire earlier hasn’t caused the scheduling of any operations to be pushed later. This is entirely expected.
- In all months but August, the weathered durations are identical to the base case. This suggests that the trenching vessel is already on hire sufficiently early and that we don’t gain much by taking it sooner.
If we look at taking the vessel on hire slightly earlier (-3 days and -7 days) we see similar data, although the durations in July have shortened. There is not enough information available to us in these charts to fully determine what is going on though.
Let’s take a look at the other extreme. In the chart below we are considering taking the trenching vessel on hire significantly later than the cable lay vessel. Here we can see that:
- The unweathered durations have lengthened. The later hire time means that the scheduling of the trenching operations is now performed a long time after the laying of the cable.
- In some of the simulations the duration range is very small, suggesting an operation which can be performed repeatable in a given duration. It is likely, but hard to tell from these charts, that we have (almost) totally decoupled the two elements of the operation.
Finally, we should consider taking the trencher slightly later than in the base case. Here we can once again see an increase in the unweathered durations, although this is slight. We can also see that there is little change to the weathered durations. This suggests that whilst our base case is a good starting point for the hire offset it may be possible to optimise this (i.e. the trencher is waiting unnecessarily for the opportunity to work on the first cable, and that we could on hire it later).
We have noted that our base case is good, but that we think we can make improvement, however, we don’t have sufficient information in this data presentation to fully inform that decision. In the next post we’ll look at how we can obtain this information and will then go on to look at how this impacts our scheduling.