The main objective with monopile design is always optimisation for minimum overall cost. The industry often talks about potential savings of 10 to 20%, but with Matt Bristow’s more efficient design techniques, these monopile savings are already achievable. The optimisation of foundations for offshore wind farms involves the investigation and examination of almost every aspect of monopile design.
Matt Bristow has been designing foundations for offshore wind farms since 1999 and therefore has a great deal of experience in reducing foundation costs. The list below shows where potential monopile savings can be made. The list is not exhaustive, but shows that with a fully optimised design it is possible to achieve savings of up to 20% or more. If you are interested in any aspect, please do not hesitate to contact me.
Click on each theme for further details:
Analysis
- Determination of wind turbine loads. The determination of wind turbine loads is normally the realm of the wind turbine manufacturer. However, there are many occasions where it may be advantageous to estimate the turbine loads independently, e.g. preliminary design or to speed-up the design process, etc. MB has developed design tools to estimate the wind turbine loads for various generic wind turbine generators, including the all-important fatigue loads. This is often very useful for early optimisation of foundation design or before a wind turbine manufacturer has been appointed.
- Reduction in wind turbine loads. In many cases it is possible to reduce the wind turbine loads. Needless to say any reduction in the wind loads will automatically result in savings in the support structure. MB has developed a variety of strategies to reduce the wind turbine loads, including wind engineering (greater use of gust size factors and turbulence intensity) and the ability to model the whole of the pile below the mudline using non-linear supports (i.e. in a truly integrated analysis). More sophisticated methods, with a greater potential for savings, include alternative time history analyses where the wind turbine manufacturer supplies time histories of the forces and moments at hub height, rather than supplying pre-calculated loads for the bottom of the tower. All above methods can result in reduction of the wind turbine loads.
Natural frequency
- Accurate natural frequencies are required in order to maximise the energy production and minimise the fatigue loads. The natural frequency is one of the most important properties of a wind turbine and is required to determine the dynamic loads on the structure and the energy production from the wind turbine generator. However, it is a common occurrence that the natural frequency of real structures are often higher than that predicted by computer models, e.g. by 5 to 8%. This can result in loss of energy production or an increase in fatigue loads. There are many reasons for this, including material, fabrication, environmental, modelling, and geotechnical considerations. MB has compiled a complete list of these factors (42 items), due consideration of which should allow more accurate determination of natural frequency, greater optimisation of energy production, and more efficient support structures.
- Incorporation of non-linear effects. It is commonplace in offshore analyses to calculate only one natural frequency, corresponding to some assumed equivalent ground support condition. However, in reality the natural frequency will vary between low and high operating loads due to the inherent non-linearity of the soil. MB has developed the capability to automatically linearise the support conditions for any applied loading, thereby permitting separate natural frequencies to be calculated for each load case. This in turn will permit more accurate determination of energy production, etc.
Fatigue design
- Improvement of S-N curves. S-N curves are required for the fatigue design of the steelwork. Offshore wind turbines are usually governed by fatigue and therefore any improvement in the S-N curves will usually lead directly to savings in steelwork. Most savings in monopile design can actually come from use of improved S-N curves, with savings of 20% or more being easily achievable. MB is familiar with the procedures set-out in the design codes that allow the use of new S-N curves based on test data. Moreover, MB is also familiar with which characteristics of the steelwork will be favourable to fatigue (or not). In this way, improvements in S-N curves of 20% or more can be achieved on a consistent and reliable basis. Selection of the most favourable S-N curve requires detailed knowledge of the many variables involved in the plate manufacturing and monopile fabrication process. Often this means working with the fabricator to optimise the welding process for the lowest S-N curve. If material characteristics are selected straight from the mill, e.g. micro-alloying, the potential savings can be even greater. On some previous projects, potential savings of up to 30% of steelwork have been achieved. As a further note, any improvements in S-N curves will potentially be beneficial for the whole structure (not just the monopile), i.e. from the top of the turbine tower to the toe of the embedment.
- Development of bespoke S-N curves (new mathematical form). MB has developed new or alternative S-N curves based on DNV C203, BS 7608, and API RP2A. These S-N curves are literally curves, rather than the conventional one or two straight lines, and differ from the latter in that they are statistically optimised for every point along the graph. The curves have a single mathematical form, and are thus ideal for optimising the fatigue design for each detail. The alternative S-N curves have the same safety level as the codified S-N curves, but with up to 5% or more savings in steel. Uniquely sets of S-N curves can also be produced to include the mean compressive stress (e.g. for plain material around J-tube hole).
Materials
- Bespoke steel specifications. MB has been involved with the development of project specific steel specifications, rather than rely solely on extant standards. When projects are large enough, it is often economical to specify bespoke steel requirements for each project, e.g. strength, fatigue, and impact requirements. This also includes the all-important verification and testing. Designed steel specifications can produce lower cost steel compared to prescribed steels.
- Project specific steels (metallurgy). Normally the metallurgy of the steel is fixed by the steel mill long before it reaches the steel supplier. However, when order tonnages are very high (i.e. >10,000 tonnes) it is possible to working directly with the steel mill to produce project specific steels. MB has experience with use of micro-alloying, whereby the steel recipe can be adjusted to reduce costs and/or to increase a particular steel property. Known alloying agents are known to increase certain steel properties, e.g. fatigue or impact resistance, with corresponding improvement in steel performance. The latter also includes weldability and any reduced need for pre-heating. Similarly, the plate processing can be selected at the steel mill to enhance particular steel properties.
Fabrication
- Optimised plate layout versus weld volume. For every pile configuration there is an optimum plate layout versus weld volume that produces the lowest overall cost and reduced fabrication time. MB has developed spreadsheets to determine the optimum plate widths and lengths versus weld volume, etc. for each pile configuration. This includes better correlation of plate widths with changes in pile diameter and ensuring all plates sizes forming the cans are optimised for minimum wastage. Even reducing the number of circumferential welds by one is a significant saving.
- Optimised plate dimensions from steel mill. MB has experience working directly with the steel supplier to optimise the plate lengths/widths/weights to correspond to the maximum ladle weights or mother slab dimensions/weights produced by the steel mill. This can result in significant savings in wastage from offcuts, or minimise the ‘surplus ratio’ as it is called. This in turn should lead to lower steel supply costs.
- Optimised welding process. The welding process is usually only optimised for speed and economy; however, it is also possible to optimise the welding process for a particular steel property such as fatigue resistance or amenability to weld improvement. MB has experience with what factors are important in determining some particular attribute of the weldment, e.g. weld profile or avoidance of undercut. This often means working with the fabricator in order to determine the optimum weld set-up (e.g. number of wires, electrode angle, and depth of flux, etc.) and weld parameters (e.g. welding speed, current, and arc voltage) for a particular application. The outcome of an optimised welding process is welds which satisfy both design and economy, ultimately leading to savings in steel.
Pile diameter configuration
- Optimal configuration. For every site/turbine/water depth there is an optimum length/diameter/configuration where the monopile and transition piece weight are minimised. For many projects this is not always achieved and more emphasis on determining the optimum configuration for each WTG location can result in significant savings or avoiding over-heavy designs, including the deepest water locations. MB has developed extensive methodologies for the optimisation of monopile design, which involves the investigation and examination of almost every aspect of design, including optimum diameter per water depth, short versus long transition pieces, and incorporation of conical section, etc.
- Large diameter/thin walled designs. Large diameter/thin walled designs can (depending on the site) have lower overall monopile weights than typical current designs, even though the outer diameter may be larger. Large diameter/thin walled designs generally have better natural frequency, dynamic, and stiffness characteristics. This in turn often means lower fatigue loads and even more savings in steelwork. Such designs are often used where the research on pile driveability permits the use of higher maximum D/t ratios (see below), or the use of improved S-N curves results in thinner plate thicknesses.
- Maximum D/t ratio. Pile driveability generally governs the minimum wall thickness for lower half of the embedment depth. The maximum D/t ratio will depend on the driving conditions, i.e. hard or easy. The D/t ratio of the lower half of the pile generally does not influence the fatigue design or natural frequency that much. Therefore significant savings can be made if much larger D/t ratios can be permitted for the lower part of the monopile. Recent research shows that resistance to pile driving depends on both D/t ratio and absolute thickness t (actually t²). D/t may be more applicable when size of obstruction is similar size to pile, i.e. overall buckling or crushing, whilst absolute value t may be more applicable when size of obstruction is small compared to diameter of pile, e.g. boulder versus large diameter monopile.
Installation
- Variable embedment depths. MB has been involved with the development of pioneering designs where the embedment depth of each monopile can be varied on site at the time of installation. Normally the embedment depths are fixed at design stage and consequently are often longer than they need to be. Typically with variable embedment depths, the as-installed depth will depend on the actual ground conditions encountered on site, as revealed by the pile driving resistance. The use of variable embedment depths is advantageous when ground conditions are unknown or uncertain, geotechnical information is limited or missing, and/or depth of weathered rock is uncertain or variable. It is also suitable for when ground conditions are harder than expected or pile refusal is reached earlier than expected. The ability to vary the embedment depth at installation stage can potentially lead to significant reduction in total installation time and associated costs. In addition, there is also reduced fatigue damage to the monopile.
- Drill-drive operations. For sites where drill-drive operations are to be used, the use of variable embedment depths can lead to a significant reduction in the number of drill-drive operations that may be required. This is especially useful where pile refusal may be only 1 metre to 3 metres or so from target depth! Minimising the number of drill-drive operations alone can make substantial savings. Variable embedment depth designs are compatible with use of sliding J-tubes (i.e. no J-tube hole required).
Geotechnical
- General. Embedment depths of monopiles can be minimised by looking at all aspects of geotechnical and pile design. In many cases embedment depths can be reduced by 5 metres or more compared to current designs. MB has developed several strategies to improve the geotechnical design of large diameter monopiles. This includes the development of extensive displacement criteria (as a requirement of the new codes) and less conservative determination of the strength and stiffness parameters. In particular, the industry generally appears to be using very conservative stiffness parameters (e.g. epsilon-50 for clays), and consequently significant savings can be made in this area. MB has developed a methodology for the determination of the stiffness parameters from either parametric equations or geophysical surveys. The latter has the advantage in that insitu measurements tend to be the least conservative. Alternatively, MB has developed a methodology whereby the soil p-y springs (lateral resistance along pile) are determined directly from geophysical measurements (G0). In addition, MB has also developed a state-of-the-art method for determining the strength and stiffness parameters for weathered and fractured rock.
- Pile diameter effects. Large diameter monopiles can often mobilise additional soil resistances. MB has developed what is known as the ‘8-spring model’ for analysing large diameter piles where all these additional soil resistances are included. For further details click here.
- Cyclic loading of piles. Embedment depths of monopiles can only be reduced so far before the effects of cyclic loading become critical. MB has developed a new analytical method to determine the effects of cyclic loading on both axially loaded and laterally loaded piles. For further information and examples click here.