For almost two months, the Lillgrund offshore wind farm in Sweden stood idle. Work to repair the export cable which transmits the electricity from the wind farm to the substation on land had to be carried out during this time. This failure raised some important questions: should cable reliability be considered in the design stage of a wind farm? What are the financial implications of failure, and is there any way it could have been planned for?
As planning for Round Three of the UK’s offshore wind farms gets underway it is likely to see bigger offshore wind farms than ever before. At the moment wind farms are limited to a couple of hundred turbines but once Round Three leasing comes into effect it is likely the numbers of turbines on any given farm will reach into the thousands. Not only will the wind farms be larger but they will also be further offshore, which will raise a whole new list of problems for transmitting electricity back to shore.
The electrical power generated by offshore wind farms is transmitted to the national grid through a system of submarine cables. Poor reliability of this transmission system will have a detrimental impact on the profitability of a wind farm. Therefore this transmission system needs to be designed in such a way that it is resilient against individual cable failures.
One of the key differences between offshore and onshore wind farms, at the concept and design phase, is the need to consider cable failure when designing the electrical architecture. This is particularly true in terms of reliability and availability of cable routes because offshore faults will take significantly longer to locate and repair. If a submarine cables fails in service the consequences for the operability and profitability of the wind farm could be dire; especially if there are delays in securing a suitable repair vessel or if weather conditions are severe, likely during the winter months.
Loss of Cable
Having to consider cable loss in this way is a relatively new challenge for the renewables industry; a particular challenge is how to take such losses into account when conducting the initial financial analysis of a wind farm. A probabilistic, software-led approach, which aims to balance the initial capital outlay for a given cabling configuration against the benefits it produces in situations like the one that recently occurred at Lillgrund, provides the toolset to approach this problem. Early analysis using such an approach, suggests that offshore renewables operators may be underestimating the impact of low probability, high risk events like this which could have negative financial implications for these projects.
Careful consideration must be given to understanding how wind farms are to be laid out; taking into account factors as diverse as turbine wake interaction, seabed topology and geology and water depth. These design considerations produce a balance between aerodynamic interactions, and electrical, operational and maintenance costs, all of which have an impact upon the whole life costs of the site and therefore the cost of energy. A plan for a site must consider all of these elements and attempt to minimise their associated costs. An optimal solution for one design consideration may impinge upon another, so these aspects should, ideally, not be considered in isolation. A holistic approach to wind farm design is therefore required.
Existing, industry standard tools allow engineers to analyse the impact of turbine placement on the aerodynamic efficiency of the wind farm and, separately, to consider the operation maintenance cost implications of certain configurations.
Once these tools have been used to decide on an array layout, the cable layout will be determined using experience and best engineering judgement. Two broad principles are likely to inform this layout - the ease of running the cable and aiming to make individual cables as short as possible. An engineer will make these decisions based on their own analysis of a number of different possibilities. Given time constraints there are only a certain number of possibilities an engineer can assess. At the moment, given the size of wind farms the optimal configuration is relatively easy to determine and this approach is fit for purpose. However, as wind farms continue to increase in size this will become more difficult.
As the size, complexity and distance from shore of wind farms increases, the approach outlined above may begin to lose its fitness for purpose. It is likely that designers of offshore electrical architectures will have too many turbine connection points for the current approach to continue to generate optimal configurations.
The larger the wind farm the more consideration needs to be given to how prone it is to individual failures – and the more difficult this is to do using the current approach to design. Clearly the costs of installation will be higher than on land. However, the reliability and availability afforded by a design are also much more important because faults take much longer to locate and repair.
It is essential that the electrical cable systems of wind farms have high reliability – that the system has the ability to withstand unforeseen circumstances – while considering the cost implications.
Redundancy in a given electrical architecture creates resilience against individual cable failures; however, adding redundancy to a design also increases the capital expenditure required to install that design. At some point, increasing redundancy will therefore become less cost effective to the wind farm. In order to analyse and identify the optimal level of redundancy for a given architecture, it is necessary to consider the system as a whole, including all major parameters of the transmission system, including:
Put simply, designers require a method of calculating risk associated with a given design: the probability of failure multiplied by its consequences – expressed as a cost in pounds. To ascertain the level of redundancy required several plans must be worked up to figure out the lowest cost implications. Frazer-Nash has developed an approach to analysis which balances these two concerns.
A suitable software toolset can first be used to calculate the total cost associated with a given cabling configuration for, potentially, thousands of nodes. This total cost includes risk values for losses due to cable failures; these are derived using probabilistic methods to predict the cost of failure throughout the life time of the farm. Using this function the cost of a given array layout can be generated.
The second element of the toolset is an optimisation routine which runs the cost routine, explained above, for multiple, different electrical architectures. This software toolset generates these architectures iteratively and attempts to optimise the whole architecture by driving towards the lowest whole life cost for a given set of turbine placements.
When combined with an array analysis toolset this gives a powerful set of design tools allowing the user to trade-off considerations against one another to achieve the optimal overall solution. Considering the extensive phase of wind farm construction which is currently planned, this holistic tool would offer a powerful method of determining the optimal layouts for these sites.
As the capacity of offshore wind farms is ultimately anticipated to far exceed that of onshore farms in the UK, there is a clear and urgent need for this approach to be adopted.
The advent of very large offshore wind farms will drive down the cost per unit of the electricity generated, making offshore wind increasingly competitive with conventional generation. However, current design methods could leave these farms vulnerable to significantly reduced output caused by single point failures. It is therefore essential that, during the design phase of the wind farm, careful consideration is given to how much redundancy is appropriate for the electrical infrastructure of offshore arrays.
Lead image: Rodsand II offshore wind farm in Denmark, courtesy Frazer-Nash