«SYNOPSIS The DOWEC projects aims at implementation of large wind turbines in large scale wind farms. part of the DOWEC project a concepts study was ...»
The fourth design line described here adopts an integral exchange philosophy. As soon as a failure is detected an integral part of the wind turbine, which may well be a complete nacelle including blades is replaced. Repair can then take place at a location with a (better) controlled environment and provided with all necessary repair equipment
DEFINITION OF CONCEPTS
The DOWEC concepts project team decided beforehand upon the exclusion of Direct Drive machines. The main reason for it was the clear preference of the industrial partners within DOWEC not to mix up possible advanced rotor and control concepts with a choice for a power conversion concept, which was unfamiliar to them. Further it was agreed that not too many restrictions had to be taken into account beforehand. Therefore both 2 and 3 bladed machines, up wind and down wind operation as well as constant speed and variable speed were taken in consideration. Apart from the wind turbine itself also the support structure (everything below the yaw bearing) was an integral part of the development of the concepts. After a number of design considerations the concepts presented in table 2 were selected.
Table 2 DOWEC concept choices plus the Direct Drive alternative Because of the more general interest of the researchers at TU Delft participating in the concept evaluation they kept the Direct drive machine as an extra option in their evaluation, hence the addition of the last column of table 2.
The Base Line concept implies a rather straightforward active stall design; a more or less scaled up version of current
1.5 to 2 MW Danish designs. Because of the size of the machine and its control system it is considered not suitable to use a monopile as its support structure. The Advanced concept represents the current trend of manufacturers to adopt variable speed pitch control. The Robust design concept aims at reduction of components and systems. No active power control, hence no blade pitching bearings and two blades only. A monopile support structure was adopted for this concept, mainly not to abandon such support structures already in the conceptual design phase.
The Stall-teeter design is inevitable coupled with two blades, having a teeter hinge at the hub, and is the most probable candidate to adopt down wind operation, together with a truss tower. The Smart-stall concept is most probably the most advanced wind turbine design, adopting variable speed with passive pitch control towards stall. This advanced design is then, most logically, combined with the base line support structure option. The Direct Drive concept, evaluated parallel to the DOWEC project team concept choices is in fact similar in its control concept to the Advanced concept, albeit that evidently a full variable speed option, inherently coupled to the direct drive generator and its advanced power converter, is implemented in this concept.
RAMS ASSESSMENT OF THE CONCEPTS
Reliability of the concepts The reliability of the wind turbine concepts will be assessed using the reduced failure frequencies for the components given in table 1. Specific concepts will then yield different total failure frequencies. For instance, omission an inverter reduces the failure frequency by 0.14. The assessment is however not a straight forward implementation application of the failure frequencies for the components given in this table, because also the (positive or negative) effects the concept choice might have on certain component reliability levels was taken into account. An example of this more detailed level of assessment of the failure frequencies is the number of failures of the pitch mechanism. It will be larger for (positive) pitch control than for active stall (negative) pitch control, due to the higher number of anticipated pitch actions. The resulting reliability levels for the six concepts are shown in figure 2 below.
The distribution of the reliability levels over the components does however not provide information about the maintainability of the concepts, i.e. about the character and the length of the repair actions that have to be taken in order to get the wind turbine in operation again. For the calculation of the availability and the costs associated with repair actions, such information is however inevitable. Thus the associated repair actions were assessed as well with respect to the concepts defined. The result is shown in table 3 below. With respect to the required maintenance action four categories were distinguished. Category 1 implies the most severe repair action and implies the use of an external (offshore) crane. The other categories require no other equipment than available at the wind turbine.
Table 3 Distribution of yearly failure rates over the maintenance categories for the 5 DOWEC concepts plus the Direct Drive concept From table 3 it can be seen that statistically every wind turbine needs to be maintained with the use of an external crane every 5 to 7 years depending upon the concept chosen. The concepts are developed with the purpose to apply them in large-scale wind farms of typically 100 units. This means that 15 to 20 lifting action per year are requested to keep such a wind farm in operation. Since the DOWEC concept deals with 5 MW designs it must be anticipated that such a maintenance crane has do deal with significant hoisting heights (hub height is typically in the order of 100 m).
Service demand of the concepts
The targeted service demand of the DOWEC concepts is a visit once per 12 to 18 months. The service period is flexible in order to facilitate opportunity-based maintenance. With such maintenance strategy the regular service of the machines is carried out in combination with a failure that needs to be repaired. Since totalled failure frequencies are in the order of 1 failure per year such opportunity based maintenance strategy is adequate for the presently defined concepts. Still a more intensive service action has to be taken every five years in which some crucial heavy-duty components are overhauled.
Availability, O&M costs and cost of energy for the concepts
In order to make a comparison of some overall figures with respect to an offshore wind farm the availability and the energy yield of 100 unit wind farm equipped with the concepts developed so far was assessed. A preliminary study regarding the availability of the concepts, for a wind farm situated some 35 kilometres off the Dutch coast in north-west direction from the Rotterdam Harbour area, shows availability levels of 85 to 94%, depending upon the concept adopted. This study is performed using a Monte-Carlo simulation tool developed at TU Delft [6,7]. With this tool it is possible to simulate a 20-year period of wind farm operation including stochastic failure occurrence of the machines and weather conditions. It was clear that the availability level for the majority of the concepts was unacceptable (a preliminary target of 94% was set at the start of the concept study); the only exception being the Robust design. The reason for it was the anticipated accessibility level of the chosen site, which is on the average around 60% (dependent upon the season) and the fact that an external crane ship is never immediately available.
Fig. 3 Yearly O&M costs for a 100 unit wind farm equipped with the anticipated concepts The yearly O&M costs were determined using a simple expert system . This expert system comprises trend lines extracted from the Monte-Carlo simulation tool, and determines the availability and related O&M costs of a 100-unit wind farm as a function of the machine reliability, an average wind speed parameter, and the distance to shore. The result with respect to the O&M costs is shown in figure 3 The level of yearly O&M costs turns out to be about 4 to 4.5 % of the total wind farm investment, which is substantially higher than the challenging 2.4% for the Opti-OWECS design solution . Two things can be noticed directly when looking to figure 4. Indeed the Robust design is most in favour with respect to the O&M expenses, but the most striking is the contribution of category 1 to the O&M costs. More than 50% of the average yearly expenses for O&M have to be paid for lifting operations using an external crane. This has a number of reasons. At first the external cranage demand of the concepts considered, which leads to15 to 20 lifting action per year for the wind farm, and second the type of crane vessel needed to perform hoisting operations at hub height (around 100 m). Figure 5 below taken from , shows the trend lines for the day rate of offshore crane vessels. It can be seen that at a hoisting height of around 85 m there is a sudden increase in day rate with a factor of 4, and evidently all external lifting operations for the DOWEC concepts have to be performed using the expensive crane vessels.
Dayrate (kEuro/day) Hoisting height (m) Fig. 5 Trend of day rate for crane vessels as a function of hoisting height The cost of energy is usually determined with the LPC approach . By adopting the standard 20 years economic lifetime and a yearly discount rate of 5% , the LPC of all concepts for the offshore location in consideration turned out to be higher than the target value of 5.5 Euroct/kWh set in the beginning of the concept study. The calculated LPC for the Robust Design wind farm however came quite close, thanks to its fairly low investment cost, its relatively low O&M costs and its high availability. These properties ruled out its 6.5% lower potential energy yield with respect to the Base Line concept and its 9% lower potential energy yield with respect to the Advanced and the Direct Drive Concept.
The results show that the concepts as presently defined are not yet suitable for further development. Definitely not when application in large-scale offshore wind farms on rather remote locations is targeted. The total level of reliability is still insufficient, although the Monte Carlo simulation for a wind farm equipped with Robust concept wind turbines showed an acceptable availability level of 94% on a fairly remote location.
The regular demand of a large external cane vessel is killing for the O&M costs and hence for the cost of energy (LPC) for all concepts. There are several ways to get over this problem. One way is to design the wind turbines such that they can completely rely on built-in facilities for exchange of components. This requests a further development of the concepts along a true modular design line.
Another approach might be to adopt the Opti-OWECS design solution approach, which includes the purchase of special purpose O&M hardware as part of the total wind farm investment. In that case a self-propelled jack-up platform was modified to perform the required lifting actions, but simultaneously serve as a maintenance base for the offshore wind farm. In this case the concepts can be further developed along the integrated component or integral exchange design lines.
1. Presently achieved reliability levels of onshore wind turbines is insufficient for application offshore
2. An increase with a factor of two of the reliability of wind turbines can be achieved against reasonable effort and costs, by choosing highly reliable components, selection of the most reliable implementation and introducing redundancy.
3. A yearly failure frequency of around one for each turbine prevents the application of such designs for more demanding less accessible sites.
4. The DOWEC 5 MW concepts as presently defined are not adequate, for real offshore applications, with the Robust design concept as the only exception.
5. Availability levels at the design location (35 km off the Dutch North Sea coast) ranged from 85% to 94%
6. O&M costs of the DOWEC 5 MW concepts, around 4 to 4.5% of the initial investment, are beyond the level that was initially foreseen. Together with the fairly low availability levels this leads to energy costs above the targeted level of 5.5 Euroct/kWh for all concepts; the Robust design getting the closest to the target.
7. Implementation of a much more "farm like design approach" may overcome the current dilemma with respect to O&M costs and availability. A purchased special purpose-lifting vessel together with an integral exchange design line for large components might simultaneously reduce lifting costs significantly and increase the availability of the wind farm.
Part of the work presented here is performed in the project ‘DOWEC concepts’, funded by NOVEM, the Dutch Agency for Energy and Environment, and is ongoing in the DOWEC project, funded by the Dutch EET-program.
1. H.B. Hendriks e.a., DOWEC Concept Study, Evaluation of Wind Turbine Concepts for Large Scale Offshore Application, Proceedings of OWEMES, ENEA, Italy, April 2000.
2. MWEP Jahresauswertung 1996-2000, ISET, Kassel, Germany, ISBN 3-9805896- 6-8/5-X/n-n.
3. W. Eggersglüβ, Wind Energie IX-XI, Praxis-Ergebnisse 1995-1999, Landwirtschaftskammer Schleswig-Holstein, Osterrönfeld, Germany.
4. M.C. Ferguson (editor); M. Kühn; G.J.W. van Bussel; W.A.A.M. Bierbooms; T.T. Cockerill.; B. Göransson; L.A.
Harland; J.H. Vugts; R. Hes, Opti-OWECS Final Report Vol. 4: A Typical Design Solution for an Offshore Wind Energy Conversion System, Institute for Wind Energy,TU Delft, The Netherlands, 1998, ISBN 90-76468-05-2.
5. Chr. Schöntag, Optimisation of Operation and Maintenance of Offshore Wind Farms, Report IW-96-108R, Institute for Wind Energy, TU Delft, The Netherlands, November 1996
6. G.J.W. van Bussel; Chr. Schöntag, Operation and Maintenance Aspects of Large Offshore Wind farms, Proceedings of the 1997 European Wind Energy Conference, Dublin, Ireland, October 1997, pp. 272-279.
7. Chr. Schöntag ; G.J.W. van Bussel, Optimisation of Operation and Maintenance, Opti-OWECS Final Report Vol.
2: Methods Assisting the Design of Offshore Wind Energy Conversion Systems, Part B, Institute for Wind Energy, TU Delft, The Netherlands, ISBN 90-76468-03-6