NEG Micon is a global leader in wind power and has operations in over 42 countries. It has installed over 12,000 wind electric generators worldwide, approximately 20% of the installed wind power capacity worldwide.
Wind Turbines are surprisingly large, with each of the rotor blades ranging in size from less than 20 meters to as much as 40 to 50 meters in length. The technology behind the design and manufacture of these items is very advanced but the logistics of moving such large items remains difficult. Transporting them from one area to another needs careful planning. It is generally more efficient to move items of this size by sea.
SIMUL8 Corporation were asked to help with considering the logistics of moving blades from a production facility with a relatively small storage area, located on an island a few miles offshore to an on-shore storage facility. The movement of blades is undertaken by a single shallow draft vessel called a barge that can move a number of blades at a time, the number to be moved being dependent on the size of the blades.
Being located on an island has certain advantages when products are to be moved by sea. There are also some disadvantages, for example weather, tides, seasonal and environmental considerations can limit the times and duration of the window of opportunity for movements. The principal constraint to movement of blades was the effect of wind - moving objects shaped like large sails is not possible in high winds.
The challenge was to understand why the small island-based storage facility was seemingly filled to capacity. There seems to be sufficient ability to move the blades from the island and understanding what was causing the build up of stock was essential.
In order that this problem could be explored a simulation of the barge movements was created. Tide times and seasonal constraints were built in (for example the annual mating of the worms that lived on the mud flats close to the barge loading dock limited the number of allowable weekly loads) and the simulation created in sufficient detail to emulate the production of blades based on a weekly schedule. A random weather effect was also built in.
The simulation predicted that, over time, the effect of the weather and other random factors would be to reduce the number of available loading slots, effectively acting as an occasional constraint to production.
For example, if there is a run of bad weather then it is possible that a number of loading and movement opportunities within a short period of time will be missed resulting in a constraint on production, as the blade storage facility gets filled up.
A static analysis of this could be undertaken, however it would base its forecasts upon, say, 20% of movements being unavailable and could assume that these were equally spaced throughout the year. The simulation allowed weather to be based on seasonal trends and so it was possible to simulate the effect of a run of bad weather reducing movements, and the consequential effect on production.
To help with this problem a vessel loading algorithm was developed to make use of the available load spaces on the vessel most effectively. This would help to minimize stock holding in the small storage facility. A range of options was developed to allow exploration of the costs and benefits of, for example, buying a bigger vessel, using a temporary floating storage facility.
The simulation provided the ability to understand the issues and to generate and investigate options. It allowed the dynamic and random effects of the weather to be be estimated and provided a greater degree of understanding to be brought to the production planning process.