Back to the programme printer.gif Print




Delegates are invited to meet and discuss with the poster presenters in this topic directly after the session 'Advanced operation & maintenance' taking place on Thursday, 13 March 2014 at 11:15-12:45. The meet-the-authors will take place in the poster area.

Behzad Kazemtabrizi Durham University, United Kingdom
Co-authors:
Richard Neate (1) F Behzad Kazemtabrizi (1) P Evgenia Golysheva (2) Christopher Crabtree (1) Peter Matthews (1)
(1) Durham University, Science Site, United Kingdom (2) Romax Technology, Nottingham, United Kingdom

Printer friendly version: printer.gif Print

Presenter's biography

Biographies are supplied directly by presenters at EWEA 2014 and are published here unedited

Dr. Kazemtabrizi is a Lecturer in Electrical Engineering at the University of Durham, England, where he also worked as a Research Associate. He has been involved in EPSRC funded SUPERGEN-Wind consortium where he has worked mainly on reliability and operation modelling of large scale intermittent wind power integration into the power grid. His research interests are operation and maintenance (O&M) of offshore wind farms, wind power reliability, energy storage, wind resource modelling, and advanced power systems modelling (control and optimisation).

Abstract

Optimisation of far offshore wind farm operation and maintenance (O&M) strategies

Introduction

A new simulation model has been presented in this paper which is aimed at optimising incurred costs of Operation and Maintenance (O&M) of far offshore wind farms. Conventional shore or near-shore based maintenance approaches are no longer economically viable for larger wind farms, situated further offshore, and aimed at increasing the wind energy penetration levels into the grid [1]-[5]. Consequently, a series of alternative maintenance policies involving crew organisation and seasonally permanent manning options are presented in this paper in order to reduce the O&M costs and eventually reduce the cost of clean energy produced.

Approach

The wind farm modelled is assumed to be situated in Dogger Bank, 125-195 km off the east coast of Yorkshire (Figure 1) with a potential capacity of 13 GW [6] as shown in Figure (1).



A simulation model has been developed which simulates the actual operation of the wind farm and consists of three components, a Component Reliability Model which is responsible for modelling the actual operation of the wind farm in terms of random failure behaviour of turbines, a Maintenance Model which essentially models several maintenance options depending on the wind farm operational characteristics, and a Cost Model which predicts the costs of repair and overall cost of energy production for the simulated O&M activities throughout a specific simulation time span. Turbine operation is modelled at the component level (two are modelled here) as a continuous-time Markov process [7]. There exists 8 distinct states for each component inside the turbine and state transitions are sampled using a non-sequential Monte Carlo state sampling approach (See figure (2) and (3)).





Six maintenance options, shown in figure (4), have been formulated based on physical parameters of the offshore problem, focusing on the performance characteristics of existing vessel and platform options [8]-[9].



Both Vessel-based and Platform-based options are considered. Vessel-based options rely on shore-based planning and organisation of repairs and may utilise either Large Boats (300+ft plus 4-crew accommodation) and/or smaller Guard Boats depending on the failure type and maintenance requirements. Jack-Up Boats required for repairing catastrophic failure events may also be considered, however, such failure events are rare and are not considered in this study. Platform-based options take into account the possibility of housing repair crew/supplies offshore, either on a temporary or permanent basis and located on either the offshore transformer platform (which is already there as part of the offshore power transmission grid) or a separate Monotower platform for an optimal turbine access from the node. Access to turbines may either be done through air (using Helicopter) or Sea (using Boats). A daily wind/wave model is also produced for the duration of simulation based on monthly averages.

Main body of abstract

1) Experimental Design:

A preliminary experiment was used to find optimal setup conditions for the experiment. A sufficient number of simulations are needed to run in order to reach a sufficient statistical certainty in results. It was found that each simulation would need to run for a minimum of 5 years in order to allow the pseudo-random nature of results to reach steady state, and allow more subtle events (such as crew staying over and type B failures) to occur. A run of 20 weather models of 10 repetitions per model with a 5 year simulation would produce a standard deviation of £0.155/MWh, a result which would provide acceptable error. This is to be applied to each proposed maintenance node. The experiments are run for a variety of changing parameters, such as the wind farm’s distance to shore (maximum 180 km), wind farm surface area, turbine ratings and different maintenance options, namely Vessel-based or Platform-based options. In this study, a target cost of £14/MWh has been taken as acceptable threshold for O&M costs based on current estimates of energy production cost for achieving grid parity for wind power [1] and [3].

2) Simulation Results:

Figure (5) shows the cost of energy per MWh for each maintenance option listed in figure (4) for the base case scenario.



The economic performance of each maintenance option has been listed in Figure (6). Clearly, vessel-based options fare far better than platform-based options, despite none of the strategies reaching the £14/MWh threshold. Estimated levelised costs for options 2 (Guard Boat), would be £125.821, which even though is still significantly higher than the required £100/MWh requirement for grid parity, it is a dramatic improvement on current UK offshore wind farm costs which range between £150-210 per MWh [1].



The cost disparity between platform and vessel options is largely a consequence of travel costs of big boat vessels and increased node operational expenditure, which results in an average increase of £52.1 M in costs for a platform option over a vessel option for the 20-year life cycle of the farm. Performance data shows maximum operating wave height is the critical factor in determining O&M costs, with the node speed, operating expenditure and small boat launch capacity having lesser effect (Figure 7).



Moreover, there are correlations between increasing distance to shore and increasing the overall cost of energy which are mostly due to an increase in the O&M cost caused by increasing costs of chartering helicopters for crew rotations. On other hand, wind farm array surface area as well as turbine types and ratings would have effects on the cost of energy produced. Ignoring guidelines on optimal placement of wind turbines there will be an optimal wind turbine density for a catchment area size based on the maintenance capacity of the offshore node because eventually turbine numbers saturate the node maintenance capacity, limiting repairs despite increasing failures. For the turbine types in this study (3.5 and 6 MW for the base case study) and for vessel-based options, the optimum number of turbines in a 30 km2 surface area is 55 at which point the higher power density of the wind farm would cancel out most of the effects of higher failure events. Similarly, as the turbines are rated higher, the cost of energy would reduce since the turbine rating is directly related to the amount of energy produced. This would suggest that for future far offshore wind farm design scenarios, should the reliability of higher rated turbines improve or their capital cost sufficiently reduced, it is beneficial to employ higher rated turbines. In conclusion, sensitivity analysis suggests that future far offshore wind farm designs should concentrate on high density, medium (up to 8 MW) rated turbines of both reliability and turbine proximity, see figure (8).




Conclusion

In this paper several different maintenance strategies pertaining to far offshore wind farms have been investigated with the aim of identifying an optimum far offshore maintenance strategy. A study quantifying the effect of vehicle characteristics found operational robustness to be the most critical factor in vessel selection for the offshore node, (responsible for 52.3% of reductions in cost of energy amongst node options) while node operational expenditure and Failure B vehicle speed would affect cost of energy to a lesser extent. The proposed strategy is a vessel-based node, with a guard boat providing the best combination of characteristics for optimal cost of energy. This produced an O&M cost of energy of £17.615/MWh which corresponds to an estimated £125.82/MWh levelised cost. The proposed offshore maintenance strategies largely negate significant effects of increasing distance to shore, affecting cost of energy for a vessel node by only £0.131/MWh per 10km offshore. Wind farm planning should take operation and maintenance into account as factors such as turbine density and rating have a large impact on maintenance operations through node-turbine travel time and failure incidence. Accurate weather prediction is of high importance in the planning and risk assessment of far-shore maintenance policies. With harsh weather conditions weather windows are small and less frequent so must be exploited - a high accuracy prediction is necessary to avoid having highly conservative operational parameters. Given the crew stayover and shore-based crew policies are narrowly economic, negligible savings mean the risks involved do not justify inclusion into a serious operation and maintenance policy.


Learning objectives
1. Development of a comprehensive procedure for modelling and simulation of offshore wind farm operation and maintenance (O&M)
2. Understanding the mathematical process of modelling turbine failure events and simulating maintenance strategies
3. Determining alternative maintenance options for larger far offshore wind farm developments
4. Identifying decisive factors in reducing the O&M costs of larger further offshore wind farms for ultimately achieving grid parity for wind power




References
[1] Parsons Brinckerhoff, 2010, “Powering the Nation”, Technical Report
[2] Garrad Hassan, 2001, “Concerted Action on Offshore Wind Energy in Europe”, Technical Report, Delft University of Technology
[3] Nielsen J., and Strensen, J., 2011, “On Risk-based Operation and Maintenance of Offshore Wind Turbine Components”, Elsevier Reliability Engineering and System Safety, 96(1), pp.218-229
[4] Karyotakis, A., 2011, On the Optimisation of Operation and Maintenance Strategies for Offshore Wind Farms, PhD Thesis, University College London
[5] Van Bussel, G., Zaaijer, M., 2001, “Reliability, availability and maintenance aspects of large-scale offshore wind farms, a concept study”, 2001, Proceedings of MAREC, 113(1), pp. 119-126
[6] Forwind Consortium to Develop Dogger Bank Round 3 Offshore Wind Farm Zone, January 2010
[7] Billinton, R. and, Alan. R., 1996, Reliability Evaluation of Engineering Systems, Plenum Press, London
[8] Pieterman, R., et. al, 2011, “Optimisation of Maintenance Strategies for Offshore Wind Farms”, , The Offshore 2011 Conference
[9] Byon, E., and Ding, Y., 2010, “Season-dependent Condition-based Maintenance for a Wind Turbine using a Partially Observed Markov Decision Process”, IEEE Transactions on Power Systems, 25(4), pp. 1823-1834