Wind farm Siting | European Marine Observation and Data Network (EMODnet)

International and European commitments to reduce greenhouse gas emissions are driving significant changes in the energy generation mix to reduce reliance on fossil fuels. Offshore wind is seen as making an important contribution to greenhouse gas emission reduction targets and many European coastal states have developed ambitious programmes for the deployment of offshore wind farms. As with the development of any new major infrastructure, there are significant challenges in ensuring that developments are technically robust and economically viable but also environmentally sustainable. The development of new industries within the marine area also needs to respect and work with many existing activities, particularly shipping, fishing and tourism. The identification of areas that are potentially suitable for offshore wind farm development therefore presents many challenges.

The objective of the Wind Farm Siting Challenge was to:

determine the suitability of sites for development of a wind farm. All aspects should be considered - wind strength, seafloor geology, environmental impact, distance from grid, shipping lanes - even if one of the factors makes this a no-go scenario;
determine whether a floating or fixed wind farm would be more appropriate.

The sites to be analysed within this challenge were at the most appropriate points within the Arctic circle in the Norwegian Sea and the Barents Sea.

This challenge was conducted in the following steps:

Assessment 1: identifying locations with the best potential for developing offshore wind. Offshore Wind Development should be economically viable. For the Arctic Ocean Checkpoint project this means identifying the most appropriate areas to locate wind farms from an economic perspective.

Assessment 2: excluding areas too important for other stakeholders from the locations identified in the first assessment. Offshore Wind Development should have little impact on other uses – including the ecosystem of the proposed area.

This was followed by phase II in which the available data sets and the underlying shared questions are reviewed.

Main results

Results of assessment 1 and 2

Fixed wind turbines
Of the 13 blocks spread out along the Norwegian coast in a technical OWE assessment, none remained after taking other sea uses into account. Most were excluded due to their locations being in major shipping routes or marine protected areas.

Floating wind turbines
Of the 290 blocks in a technically suitable area, 124 blocks remain after taking other sea uses into account (table 1 and figure 1). Six of these blocks are in the Russian part of the Barents Sea (ICES area Ib), on the Murman Rise. The remaining blocks are in Norwegian waters (ICES area IIa2), mostly around the Lofoten and Tromsø. West of Trondheim the combination of other sea uses results in only a few remaining OWE blocks.

Table 1. Main characteristics of the remaining blocks with potential for developing (floating) offshore wind energy parks in the Norwegian Sea and Barents Sea, based on the first round of assessments within ‘SeaBasin Checkpoints – lot Arctic’.

ICES area	No. OWE blocks	Mean distance to port (km)	Mean water depth (m)	Mean wind speed (m/s)	Area (km2)
Ib	6	238	179	8.2	1553
IIa2	118	203	272	8.4	29969

Figure 1: Map showing the area with potential for developing (floating) offshore wind energy parks in the Norwegian Sea and Barents Sea, based on the first round of assessments within ‘SeaBasin Checkpoints – lot Arctic’. Also shown are human settlements that may play a role in this development.

More details on the approach and detailed results can be found here:

1st Wind Farm Siting Challenge - Assessment part 1

To have determined the best technical and economic area several data sets were required to provide information about the options to develop offshore wind turbines. Data sets were searched to support the following objectives:

An economical and viable development needs a market for the electricity it produces. This means the presence of cities, ports or large industrial sites. Alternatively, high voltage connection points can also be used to transport the electricity to a market. Currently existing connection points are often located near cities, ports, large industrial sites and power stations (conventional/nuclear/etc.);
Developing offshore wind energy also requires facilities like ports, quays and cranes to support the activities of building, operating and maintaining the turbines and the supporting infrastructure of cables and transformer platforms. This also means the necessary on-shore availability of motorways, railways and airports, e.g. to allow specialized persons or replacement parts quick access to the area.

Locations that satisfy objectives 1 and 2 are presented in Table 1 and figure 1.

Table 1 Locations representing both the presence of a market and infrastructure available for offshore wind energy development (data from a.o. Wikipedia and other websites of municipalities and ports). Suitability or presence of infrastructure was indicated as followed: 0 absent/unsuitable; 1 present/suitable; 2 with limitations.

Name	Inhabitants	Port Maint	Port Cons	Heavy Ind	Grid Connect	Railroad	Motorway	Airport	Country
Murmansk	300000	1	1	1	1	1	1	1	RU
Severomorsk	50000	1	1	0	1	1	1	1	RU
Tromsø	70000	1	1	0	1	0	2	1	NO
Bodø	50000	1	0	0	2	1	2	1	NO
Trondheim	180000	1	1	0	1	1	1	1	NO
Narvik	18000	1	1	0	1	1	2	2	NO
Hammarfest	7500	1	0	0	1	0	2	2	NO
Kirkenes	3500	1	1	0	1	0	2	2	NO
Longyearbyen	2000	0	0	0	0	0	0	1	NO
Nikel	12500	0	0	1	1	0	2	0	RU
Archangelsk	350000	1	1	0	1	1	2	1	RU
Severodvinsk	190000	1	1	1	1	1	2	2	RU

Map of the Norwegian Sea and Barents Sea

Figure 1: Map of the Norwegian Sea and Barents Sea showing the locations from the table above, based on assessment 1 and 2. Yellow-to-orange colours indicates the identified area was suitable for offshore wind farm development, darker colours signify higher mean wind strength.

Once the locations with suitable market and infrastructure were identified, two further data sets became important:

Bathymetry (water depth, metres) (GEBCO 2014 gridded bathymetry 0.0083 degrees)
This indicates locations with suitable water depths that are compatible with either a fixed or a floating offshore wind turbine.
Wind strength (m/s) (Copernicus Marine Environmental Monitoring Services or CMEMS)

CMEMS had a set of satellite-derived data sets (with global coverage 0.25 degrees) available covering six recent years of monthly wind climatology. Data was available for a total of 59 of the 72 months of the six year period. To ease the analysis an average wind resource was calculated and applied.
The data collecting instrument on the satellite, a scatterometer, came with its own limitations. One of these was that a scatterometer cannot determine wind speeds over land or over an ice-covered sea. Thus the average wind resource dataset had no data where sea ice had prevented observations during the observation period. However this was not a severe problem given the current technology.

There was sufficient data available for the wind resource dataset. The available wind resource for the Norwegian Sea and, where data is available, the Barents Sea, was comparable and possibly somewhat larger than that for the North Sea. This dataset also provided the resolution used in the analysis. By combining the market and infrastructure data sets with coastline geography data sets (Arctic countries coastlines), a third important dataset was developed, Distance to port and market
This was useful to determine how far it was possible to develop an offshore wind farm. This distance was an important economical factor as it determines many costs, such as length of the HV-cables that brought the generated power to the market, and the travel time required when building, operating and maintaining the wind farm.

Assessment Part 1
To properly interpret the data sets outlined above it was necessary to make some realistic assumptions about offshore wind technology options: .

How far from port and market it was economic to attempt to develop OWE.
The maximum distance within the North Sea, currently one of the best developed OWE areas, is a little over 200 km (close to the centre, on the Doggerbank). For the purpose of this study the upper limit was set at 250 km.
How water depth was likely to interact with offshore wind turbine technology.
1. Fixed OWT can be built to water depths to about 50 m deep, with several construction options including monopoles, tripod, jacket and gravity-based (concrete) foundations structures. This choice more or less follows the current state of technology in the North Sea
2. Floating turbines can be built in water depths starting from ca. 100 m. to ca. 500 m. deep. These numbers match with pilot installations and early wind farm developments for turbines such as the HyWind (SPAR-type) and WindFloat (floating jacket). This technology was still in its infancy and not many such turbines were in operation.
  Data assembled by the EU-project ACCESS on a.o. SPAR-type platforms (D 4.21) shows that these are reasonable assumptions

Sea ice and offshore wind turbines
Offshore wind parks had not yet been built in (sub-)Arctic waters, where sea ice (or larger ice floes) can occur. Technically it was possible to build wind turbines strong enough to withstand the forces exerted by sea ice. It increased their cost for which the economics of offshore wind turbines do not allow. Icing of the turbines blades can form one more complication that requires both technical attention and increases costs of maintenance. The pragmatic choice made for this study to not suggest OWE development in ice covered waters due to the lack of satellite wind speed data over ice covered sea areas was therefore also justified by economics.

While technology options to construct ice-resisting OWT did exist (see reports of ACCESS), the increased costs made them not currently viable. This is an active field of research within wind turbine development.

1st Wind Farm Siting Challenge - Assessment part 2

The second step of the assessment is to exclude those areas where an impact on the ecosystem or other uses of the sea area can be expected from the technical and economic potential determined in the first step. For this study four marine uses have been considered:

Fisheries, using a dataset generated for OSPAR by ICES Working Group on Spatial Fisheries Distributions (WGSFD). Data included information on the surface interaction of mobile bottom-touching fishing gears (Swept Area Ration or SAR). Fishing activities do not combine well with offshore wind turbines as the fishing gear may cause damage to electricity or mooring cables, and the presences of wind turbines may also pose a danger to the fishing vessel and its crew. However, other types of fishery can be performed within an OWF, e.g. angling and other passive gears.
Shipping, using a dataset provided by Halpern et al. (2015). The dataset has global cover, is in high detail (ca. 1 x 1 km) and of a recent year (2013).
Marine Protected Areas, to respect nature conservation areas the MPA-database that was assembled for the MPA Challenge was re-used. For the Wind Farm Siting it was augmented with a few Norwegian MPA that were not included in the original dataset (too far south), but were relevant at this point. The MPA-database relies heavily on the WDPA, but many other (national) sources were reviewed, leading to a few dozen additions across the whole Arctic.
The current assessment for potential OWE sites was based on known protected locations within our current vision because assessing every location in the wealth of detailed data available, especially for Norwegian waters, was beyond the scope of this project.
The MPA-dataset serves here as a proxy to ensure minimal impact on the ecosystem.
Oil and Gas infrastructure, including pipelines (Norwegian Petroleum Directorate or Oljedirektoratet). These were incorporated using the map services offered by the NPD to see the positions of oil and gas infrastructure (platforms and subsea installations) as well as pipelines.

When implementing each of the three above data sets, suitable cut-off values or choices were required. For the fourth dataset the mere presence of oil and gas infrastructure was seen as sufficient.

For fisheries, the effort within the technical OWE area was considered. Due to a lack of pattern, the final selection was made manually. OWE-blocks with several (five or more) and relatively intense recorded fisheries activities were excluded from the locations selected for potential OWE development. Access to fishing grounds that could become isolated within wind farm areas were also considered. Within the technical OWE area, 75% of the fishing effort was to be respected and this area was excluded from the potential OWE development sites.

For shipping, connectivity and access to ports is vital. Locations with 600 or more ‘vessel movements’ were considered too important for the shipping industry and were therefore excluded from potential OWE sites.

For Marine Protected Areas (nature conservation and general biodiversity considerations), the existing MPA-database was used in combination with a 5 km buffer distance outside. OWE development must therefore be located at least 5 km away from nature conservation areas, as a number of the MPA are also home to bird nesting colonies. Thus ensuring a free corridor in and out of MPAs, unobstructed by wind turbines for foraging and or migrating species.

For Oil and Gas infrastructure, the presence of any infrastructure has led to excluding the blocks from the technical OWE area. The selection had to be made manually as the software did not adequately detect overlap.

Very few OWE-blocks were excluded due to a single obstruction. In most cases shipping obstructions combined with either fishing grounds or areas in or near MPA.

Detailed maps of assessment 2

Map showing OWE development

Figure 1: Combined map showing OWE development potential for (floating) offshore wind turbines in the Norwegian Sea and Barents Sea, as well as dropped area with an indication of the other sea users that were given precedence.

Maps for competing uses 1

Figure 2: Map showing the available information on fishery effort in the Norwegian Sea and Barents Sea and surrounding areas.

Maps for competing uses 2

Figure 3: Map showing the available information on shipping in the Norwegian Sea and Barents Sea and surround areas.

Maps for competing uses 3

Figure 4: Map showing the available information on MPAs in the Norwegian Sea and Barents Sea and surround areas.

Maps for uses 4

Figure 5: Map showing the available information on oil and gas facilities in the Norwegian Sea and Barents Sea and surround areas.

Results of phase II

The Arctic Checkpoint – Wind Farm Siting project derived its data sets predominantly from sources outside EMODnet. This is mostly due to the fact that the study area is located outside the focal area of EMODnet, and is therefore not covered. This situation may change in the future as the Arctic has been recognised as an area where more attention from the European Union, and therefore also from EMODnet, is warranted. The main dataset for this challenge, the wind resource, was available from Copernicus and thus from an EU-related source.
Two ecosystem-related data layers that were included in the plans could not be included. No data sets were found that could be used as a reliable basis for 1) bird migration routes and 2) sea mammal migration routes. This should be labelled as an identified data gap. It is however not necessarily a data gap that is specific to the Arctic. Such maps/data sets are also not be available for e.g. the North Sea.

More details can be found here:

Wind Farm Siting Challenge - Phase II

For Phase II the focus for the Wind Farm Siting Challenge was on reviewing the available data sets and the underlying shared questions of the Checkpoint Projects with the challenges as tools to assess and clarify the points below. It included the identification of:

synergies between different monitoring, observation and data collection programmes
how well the current data meets the needs of users
knowledge gaps
views where new technologies will allow for improvements (quicker, more accurate)
which temporal and/or spatial resolution of data products are required
prioritising the creation of new data and improving the availability and usability of existing data.

Confidence
Confidence in the information available is an important issue. This consists of two different aspects: one statistical and one broader, more subjective aspect with regards to selected data sets and assumptions.

Statistical confidence and confidence limits can be calculated, and included in tables and potentially as additional data layers in (interactive) maps. The wind resource dataset can be utilised in this way as variance and other statistics can be calculated across the available years.
While the bathymetry dataset could also be used in this way, it would require access to the underlying data sets to assess variability in the measurements (e.g. in different locations within a ‘grid cell’).

Statistical measures of confidence could also be applied to the data layer on waves. Copernicus widened their catalogue so that wave data covering (most of) the study area of the Wind Farm Siting Challenge is available. However, the current service release offers real-time modelled wave data and in-situ observations. For the purpose of this challenge, a reprocessed product giving e.g. monthly means calculated across several would be more useful. Such a product is included in the product ‘Roadmap for Waves’ for a release in April 2018. Statistics on waves, significant heights, extremes etcetera are useful when calculating the required strengths of engineered constructions such as foundations for fixed OWT, and moorings as wells as bodies for floating OWT.

The fourth data layer where confidence could be included is biodiversity as a measure of the environmental intensity of use on which OWT should have minimal impact. For pragmatic reasons biodiversity was based on the MPA-dataset, which was already being collated for the SBC Arctic challenge on Marine Protected Areas in this project. The MPA dataset may show varying confidence levels and initiate questions about the underlying data sets and selection and designation processes. Are the MPA themselves in the right location? Are they of the right size? These questions are beyond the scope of the wind farm siting challenge, but to some extent addressed within the MPA challenge with an analysis on coherence of the MPAs as a network.

Confidence is also important in the activity data layers for shipping and fisheries. For example: how accurate are the available data sets in sufficiently representing the actual activities within the area? To the best of our knowledge the selected data sets represened the best available data that consistently covers the study area. Smaller, more localised but also more detailed data sets could be used, but these would complicate the analysis and may not improve confidence as they involve different collection and processing methodologies. For example, the locations and boundaries are official and precise for shipping lanes, cables, pipelines and oil and gas platforms, and to some extent MPAs. Therefore there is a high level of confidence for these activities.

Marine Spatial Planning
MSP is the main focus of the Wind Farm Siting Challenge according to the questions posed for the challenge in the tender document. As such the focus has been to find areas that are best suited for developing offshore wind farms while at the same time respecting the presence and needs of several other (competing) users of the marine environment.

Due to the scope of the project we did not attempt to gauge the cost of constructing the wind farms, the amount of kWh produced annually, nor what changes to design parameters such as hub height, blade length would make e.g. with respect to vertical wind profiles.

New and improved data sets
Since our initial assessment (Phase I) some of the data sets have been improved and a few new ones have come to our attention. We present these new developments and discuss them, mainly to estimate what this would mean if the analysis was repeated.

Wind: Copernicus - the dataset has been updated and now has data for more years (2012 to 2016 have been added). No other developments.

Wind: A search was made for an alternative wind resource dataset, and an annual mean wind speed dataset (at 10 m height) was identified. The horizontal resolution is stated as 12 km for sea areas. This dataset could be an improvement on the global scatterometer based dataset available through Copernicus. However, it is only available as a WMS which, although valuable for visualising, does not allow (re-)use of data in other GIS operations. Contact data for the Norwegian Meteorological Institute is provided so more options may be available on request. Related data sets are available for ice concentration and wave height (significant) from the same source.

wind speed WMS Norwegian Meteorologish institutt
Figure 1: Zoomed view of annual mean wind speed from WMS by the Norwegian Meteorologish institutt served by GeoNorge as part of NorgesKart.

Waves: new from Copernicus

Can be used as engineering information. E.g. to avoid for OWE development areas with an extreme and thus costly wave climate.
By avoiding these areas for OWE development, they are also left available for the potential development of wave energy as a renewable energy source. However, wave energy may also be better off with not the most extreme wave locations, but with a fairly mild and predictable and harvestable wave resource.
Starting with the service release of April 2017, the available data includes real time, in-situ observations and real time models. For the purpose of the Wind Farm Siting Challenge a reprocessed dataset e.g. a wave climatology with monthly means, extremes, etc. would be more suitable. In the product roadmap such information may become available as part of the release that is planned for April 2018.

Fisheries: JRC – MFA

The EU-JRC has reported an interesting development regarding the use of AIS to map fishing activity. The ‘Mapping Fishing Activitiy’ or MFA tool is presented within their EU Science Hub, more specifically as part of the BlueHub and a research article by Vespe et al. (2016) has also been published. The WebGIS tool delivers the results as a WMS, as shown in Figure 2. An email address is offered, to contact for download access.

Bluehub webgis MFA results

Figure 2: BlueHub WebGIS showing the MFA results (blue colours, contrasting edge) and data coverage (red colours).

Vespe, M., Gibin, M., Alessandrini, A., Natale, F., Osio, G. C., Vespe, M., … Osio, G. C. (2016). Mapping EU fishing activities using ship tracking data Mapping EU fishing activities using ship tracking data. Journal of Maps, 5647 (April 2017). http://doi.org/10.1080/17445647.2016.1195299

Oil and Gas Platforms: EMODnet Human Activities portal
The Human Activities portal from EMODnet has seen considerable improvement over the last year. However as the study area for the Arctic Wind Farm Siting Challenge is mostly outside the focal area of the portal, many data sets are not relevant. However, the Oil and Gas Platforms dataset available is relevant and it presently covers the full set of Norwegian platforms in both the Norwegian Sea and Barents Sea. It is now a viable alternative to the authoritative data available from the Norwegian Petroleum Directorate, as used in Phase I.

Bathymetry: EMODnet Bathymetry portal
The Bathymetry portal from EMODnet has updated data that is of some relevance to the Wind Farm Siting Challenge. The geographic cover has been widened so that the Norwegian and Barents Sea, also including the Murmans Rise in the Russian part. This dataset offers improved horizontal resolution 1/8 * 1/8 minutes, where the GEBCO2014 bathymetry used in Phase I has ½ * ½ minutes.

FINAL Assessment Wind Farm Challenge, Phase II
Despite the availability of a few improved data sets, the project team felt that a complete re-analysis of these challenges was not warranted. The main dataset on wind resources is mostly unchanged and the outcome would be only a minor revision. The resources that are made available, will be redirected towards the MPA challenge where further effort is needed. Similar advice on repeating the challenge were obtained from the Stakeholder Meeting and the Expert Panel Meeting.

A report prepared by a group of Norwegian authorities led by the Norwegian Water Resources and Energy Directorate “Havvind, Forslag til utredningsområder” (NVE, 2010) was identified. The 15 areas that have been studied there are roughly located in the same areas as our results, though mostly closer to shore.

Another on-going project that may provide helpful information for a.o. Wind farm Siting is NOVASARC, Nordic Project On Vulnerable Marine Ecosystems And Anthropogenic Activities In Arctic and Sub-Arctic Waters (http://novasarc.hafogvatn.is/). A suggestion for an interesting and different approach was received from one of the experts following the SBC Arctic workshop; the Svalbard case (June 2017) for ConSite, a consensus-based siting tool aiming to make the process of siting a new development more democratic, verifiable and cost-effective (http://www.nina.no/consite).

NVE. (2010). Havvind, Forslag til utredningsområder. Oslo. Retrieved from https://www.regjeringen.no/globalassets/upload/oed/rapporter/havvind_ver02.pdf?id=2181946

Problems and gaps

The data sets that were used are available on the internet, but these should be easier to find and more readily available. Discoverability is often low.

Lessons learned
The final results suggest that with the available data an adequate assessment can be made on the potential development of offshore wind parks in the Norwegian Sea and Barents Sea.

Offshore development of wind energy in this region will have to rely on floating turbine technology. This technology may need several more years to mature sufficiently before successful deployment in Arctic and sub-Arctic waters. For an in-depth assessment of the economics of an offshore wind farm, specific information will be needed regarding the technology used for determining moorings and assessing geophysical conditions on and in the seabed.

Recommendations
For the next round of assessments, a smaller and more detailed block size for the OWE-analysis could be used. This could result in less area being excluded due to other users and/or ecosystem concerns. The wind resource data, at a resolution of 0.25 degrees, was not detailed enough to more precisely define geographical boundaries for development. However as this resource is mostly changing gradually in a repeated assessment a smaller block size can be used.

Now that the three most promising areas are identified, the possibility exists that more suitable data sets can be found. Data sets that do not cover all of the original study areas, but do cover at least one or two areas can be used, although preferably all three areas would be included in one dataset.

References and Links: