Offshore wind farms plan for deeper waters
Battered by wind and waves above the ocean’s surface, offshore wind foundations can’t take chances on what lies on the seabed.
Lying 53 miles off the coast of the Netherlands, the Gemini wind farm began running at full capacity in May this year, making it the world’s second largest wind farm. Built in just two years, the wind farm narrowly missed out on beating the London Array in the UK’s Thames estuary in terms of generating capacity – 600MW versus 630MW. But the farm is in waters that are 12m deeper than those of the London Array so that it could lie out of sight of the coast and pick up the stronger winds of the North Sea.
As wind farms move further offshore, their foundations become a potentially costly part of the design. To pick up on the stronger winds of the Atlantic Ocean, wind turbines will need to move into much deeper waters than the 25m of a farm like the London Array. Not only do they have to withstand bigger stresses from waves and the winds themselves, the cost can increase dramatically. In existing arrays that sit in relatively shallow waters, foundations cost as much as a third of the total for the entire turbine. In 2013, RWE dropped plans to build a 1.2GW array in the Bristol Channel having spent £13m on studies for the project. The German utility cited the 45m depth of the water there and “incredibly challenging” seabed conditions as among the reasons for cancellation.
Builders of offshore wind turbines are not short of choices for foundations. The simplest is the gravity base and was used on the first offshore wind farm at Vindeby in the shallow waters amid the islands of eastern Denmark. The gravity base is simple: it is formed from large concrete slabs that are shipped out to the site, lowered onto the sea floor and then held down with extra ballast. Because they simply rest on the surface, the ability of the mast built on top to resist movement depends heavily on the strength and behaviour of the soil as well as the weight of the structure itself. Traditionally, because of the amount of metal and ballast they need, gravity bases were used in oil rigs operated in shallower waters and then only in ‘one off’ conditions, according to Erik-Jan de Ridder, senior project manager at Dutch consultancy MARIN.
Today, a more common option than the gravity base is the suction caisson: developed for situations where gravity cannot provide enough force on its own to build a stable foundation. When pushed into the soft seabed, possibly with the aid of air and liquid being pumped out of the caisson’s skirt, the assembly is sucked into the ground. Wind farm operators now use the suction caisson for relatively shallow waters. The Carbon Trust picked a design concept by SPT Offshore in a competition and sponsored DONG Energy to build the suction-bucket jacket for use in a demonstrator based on a Siemens turbine lowered into 25m of water off the coast of Germany in 2014.
Another option is to pound piles deep into the soil using drivers sitting on a rig that is itself temporarily secured to the seabed. One method is to sink a single large pile into the ground and place the mast on top, but to save weight and cost, some use jackets of steel latticework that either surround a mast that lies half underwater or provide a platform for the mast above sea level. The open lattice structure of the jacket reduces the amount of metal and other materials needed for the foundations. Contractor Ramboll opted for a pile-driven monopile design for the Gemini wind farm.
The use of piles calls for a stable rig to be in place to drive them into the seabed. Norway-based OWEC, for example, uses a rig to first sink a piling frame that provides the template for the piles. The rig hammers the piles into the ground before lowering the jacket onto the piles and cementing them in place.
De Ridder says there is increasing potential for gravity bases to be used on deep-water masts because, without their ballast onboard, they can be moved to the target site by tugs and lowered into place. They do not need the large crane vessels of the pile-driving techniques.
Making sure the mast lasts for more than a couple of decades is not just about making a mast heavy enough or driving a pile far enough into the seabed to stop the whole structure from moving around or tilting under the pressure of wind and water currents. The combination of mast, foundation and surrounding soil changes the critical frequency of resonance of the entire structure. Get the calculations wrong and the mast could become dangerously unstable. In the Gemini wind farms, the diameter of each monopile depends on the depth and type of ground into which its piles are sunk.
Mast designers worry about two frequencies. It would be easier if they were fixed frequencies but even that option is going away. To make it easier to connect turbines to the AC grid, older turbines used synchronous generators with the blades rotating at a constant speed no matter the wind conditions. This rotation leads to a pair of critical resonant frequencies in mast design. The first, known as 1P, results from vibration at the hub caused by imbalances in the overall rotor array.
Then there is blade shadowing. As each blade passes in front of the tower, it reduces the wind load on the mast itself. Naturally, the frequency of this is the number of blades multiplied by the rotation frequency. So, it’s 3P for the three-bladed generators found in most offshore farms.
A further source of stress comes from the waves and currents that surge around the mast, with frequencies of 0.1Hz or less, leaving a gap between that and the 0.12-0.25Hz or so of the 1P band and 0.35-0.75Hz of the 3P. The big problem for wind turbine designers is the relatively narrow slots that sit between the 1P and 3P vibration frequency ranges. One option is to aim for a structure where the mast’s resonant frequency is above 3P. Unfortunately, that points to wide, heavy towers that become prohibitively costly. Realistic designs need to employ mechanical design that pushes their natural frequency into a range that sits between the 1P and 3P bands.
Today, wind-farm operators are turning to high-voltage DC transmission back to land because of its higher overall efficiency and ease of integration compared to conventional 50Hz or 60Hz AC. As the speed of the turbine does not need to be linked to the grid frequency, they can use variable-speed designs. And the resonant 1P and 3P frequencies become ranges to the point that there is no clear gap between the top of one and the bottom of the other. Finding a safe resonant frequency is now even harder. And the seabed does not make life any easier.
Normally, the assumption is that the flexing of the mast as it is battered by the elements will loosen the ground around the foundations and make the overall structure less stiff. The design policy is to account for that and assume the resonant frequency will fall over time. So, the best option seems to be to err towards the 3P zone. Experiments with centrifuges and measurements of existing structures have found the opposite can happen: the sand and soil locks into place and makes the entire structure stiffer, pushing it into the 3P band.
For example, according to work performed by Professor Stuart Haigh of the University of Cambridge, soft clays tend to swell as soil works loose over time and the stiffness falls. Monopiles in sand, on the other hand, will shift in such a way that they compact the ground around them. The soil falls into cavities as the monopile leans in one direction. When the mast leans back the pressure on soil that has nowhere to go compacts it into a hardened mass.
It’s not easy to simulate the stresses of offshore foundations to find out how soils react under the enormous weight of turbine masts. Last autumn, Professor Mark Cassidy’s lab at the University of Western Australia’s Centre for Offshore Foundation Systems installed a centrifuge big and powerful enough to spin a bucket holding two tonnes of soil at three revolutions per second.
Speaking at the recent Lloyds Register Foundation conference in London, Cassidy said of scaled-down models in a bucket of soil: “The stresses are all wrong.” More extreme measures are needed: “What we do is accelerate to greatly increase the unit weight and so get the correct stress relationship between the large offshore foundation and the much smaller physical model. So we spin soil at up to 200g and then we can get the one to two hundred scaling we need.
“We also look at numerical modelling techniques to put this into a theoretical context. The idea is that we can come up with models or analytical formulas that engineers can use in their daily design.”
Another possibility is to avoid the problem of foundations in the first place and simply build the turbines on floating rafts. Manufacturers could assemble them in dock and have tugs simply tow them out to their final destination to be ballasted and anchored in place.
In June, the Carbon Trust moved to the second phase of a series of studies into the commercial viability of floating wind platforms and whether it will be feasible to unload their ballast and tow them back to port for major repairs.
Floating a wind turbine does not make it necessarily cheaper to build than today’s fixed structures – and the work to model the seabed still matters. Because the technology is still in its pilot stages, the true costs are unclear but some researchers believe that up to 35 per cent of the money could still go into the anchoring systems.
Many of the anchoring systems use the same core techniques as those employed for fixed structures, such as suction caissons and driven piles, with the added complexity of deployment in depths of 40m or more. Dragged anchors are cheaper to use but can only be used in certain types of soil, such as soft clays. They also carry a lot more risk of coming unstuck and damaging the seabed as they move.
Brian Diaz and colleagues from Texas A&M University, working with the universities of Amherst and Maine, argue that one option for cutting costs is to share anchors among the turbines in a large farm using a hexagonal grid. This reduces the number of anchors an individual turbine needs but also provides some redundancy for each turbine if a cable snaps.
For the UK studies, the Carbon Trust has charged consultancy Ramboll, which designed the foundations for the Gemini farm, with the job of analysing the costs and risks of different mooring systems for floating wind platforms. Although the costs might not come down with floating wind turbines, they provide the opportunity to harvest more energy through hybrid designs. Denmark’s Floating Power Plant developed a wave-energy harvester that can support a wind turbine on top of its platform, which uses a semi-submersible design borrowed from the oil and gas industry, according to UK general manager Chris McConville.
Because the Danish waters are shallow, the company set up the UK subsidiary to commercialise its technology, with pilot projects planned off the northern coast near Caithness and south-west of Milford Haven.
Irish consultancy JJ Campbell & Associates aims to assemble hollow concrete caissons into a V-shaped wave harvester and wind-turbine platform. Water surging into the bottom of each caisson creates pressure cycles that drive turbines in a similar way to pistons in a car engine. Wind turbines stand on a steel superstructure built over the top of the caissons. According to JJ Campbell civil engineer Emmett Farrell, the aim is to build the structure onshore, float it out to 50km or so off the coast and moor it with a gravity base.
An advantage of wave-harvesting platforms is that by removing energy from the ocean’s movement, they create stable harbours in the lee of the structure, which can be used to improve access to the turbine array for maintenance. US-based Float Inc aims to turn floating wind and wave generators into transit harbours for container ships. The offshore ports would transfer containers carried on massive carriers to smaller ships destined for local ports.
The pressure is on the technology of foundation design to develop cheaper ways of keeping energy harvesters stable at sea. The hybrid options may provide the final push to make the plans viable as they head into deeper waters.
Keep the noise down
If you want to secure an oil rig or wind turbine, piles driven deep into the seabed look to be one of the most reliable options. But the sound of hammering pile drivers can hit 180dB and carries for miles underwater – disturbing whales and other sea creatures with sensitive hearing.
Today, to limit the noise to an only slightly less punishing 160dB, crews use curtains of air bubbles. But for two teams working independently, the answer is to put down the hammer and use a different tool. In The Netherlands, engineers at GBM Works won the Philips Innovation Award in 2016 for their plan to liquefy the soil around the pile temporarily so that it can be driven in using far less force. The liquefaction process helps fix suction caissons in place.
In the UK, an EPSRC-funded project headed by researchers working at the University of Dundee is proposing to use a screwdriver instead. The screw piles are made using a series of plates welded into a helix around the outside of the core cylinder. They could, in principle, be unscrewed out of the hole during decommissioning.
To test the approach, the researchers constructed 1:50 scale models and used a robotic design to screw them into dense sand while being spun in a centrifuge to create a force of up to 50gf. The researchers aim to incorporate the results into the numerical-analysis software used by pile designers.