Modeling Offshore Wind Installation Costs

Abstract

The costs of offshore wind power exceed onshore wind costs, in part due to added expenses associated with offshore installation. In this chapter, we model the installation costs of offshore wind projects and perform sensitivity analysis to identify the variables most responsible for uncertainty and risk. We adopt a bottom-up engineering approach based on current technologies and expected market conditions for the period 2012–2017 to estimate stage-specific installation costs. The time required for specific installation activities (travel time, loading time, installation time, etc.) is estimated based on empirical data and benchmark studies described in previous chapters. Offshore wind farms are characterized in terms of their nameplate capacity, turbine capacity, and distance to port and shore. The user is required to provide information on vessel selection, installation strategy, and market and contract conditions, and the output is the cost as a function of the model factors. The model is illustrated for three planned US wind farms (Cape Wind, Bluewater Delaware, and Coastal Point Galveston) and costs are estimated to range from $130,000 to $370,000 per MW. Cost is relatively insensitive to distance to port, but unit costs decline significantly with larger turbine capacity and increase with the time required for installation. The limitations of the analysis are described.

A. Parameter Verification

9.1.1 Temporal Parameters

Empirical data on the unit time to install turbines and foundations at existing European wind farms are compared against the weather-adjusted time per unit (ADJUNIT) output from the model to confirm the parameterization. In Table 9.17 the output of a range of assumptions from Table 9.7 are used to estimate the per foundation installation time. In Table 9.18, similar assumptions from Table 9.9 are used to determine the per turbine installation time. The parameters were chosen to give a complete range of installation times and the empirically estimated times for foundation estimation range from 1.3–5.5 days with an average of 3.6 days, and for turbine installation from 1.3–9.5 days with an average of 4.0 days. In both cases, the model output is close to the mean empirical value and the range of values generated by the model matches closely with the range of values in the empirical data.

Table 9.17 Alternative parameterizations impacting foundation installation time and output time per foundation

Table 9.18 Alternative parameterizations impacting turbine installation time and output time per turbine

9.1.2 Output Costs

Installation contracts are proprietary and are generally not publicly available, however, in three recent UK projects for turbine installation using SPIVs (Walney, Sheringham Shoal, and Greater Gabbard), cost data were made public and is used as a check on parameterization. In cases where the model output deviates from reported data, adjustments to the model parameters to account for known factors yield reasonably accurate cost. This illustrates the unique nature of each project and the need to account for individual project attributes through alternative parameterizations or adjustment factors.

Walney

Seajacks signed a $70 million turbine installation contract at the Walney windfarm [14]. Walney is composed of 102, 3.6 MW turbines and is 40 nm from port. The effective contract dayrate is $130,000 per day. When these data are input into the model (assuming expected values when not otherwise provided), the average turbine installation time is 3.2 days, the total installation time is approximately 11 months and the estimated cost is $45.9 million. The difference in cost between the model output and the contract is due to the installation time discrepancy. While both the model and the contract have a similar turbine installation time (1–4 per week in the contract and 2.2 per week in the model), the model gives a total work time of 11 months while the contract installation time is 18 months. This discrepancy is due to expected high weather downtime through the winter [15]. When the weather factor is set at 50%, the estimated cost is $74 million and the total work duration is 17.5 months, consistent with the published contract cost and duration. While there will be weather delays for U.S. projects, they may not be as severe as those experienced in Europe.

Sheringham Shoal

Master Marine signed a €78 million (approximately $101 million using 2009 exchange rates) contract for turbine installation at the Sheringham Shoal wind farm. Sheringham Shoal comprises 88, 3.6 MW turbines and is 20 nm from port. The effective contract dayrate is $374,000 per day. Using the default parameters yields an output of $37 million, however, when the dayrate is changed to reflect the contract dayrate, the model output increases to $104 million. This suggests that while the model dayrate is too low in this case, the temporal parameters are reasonable. The model dayrate is likely too low because of the size and capability of Master Marine’s vessel and high competition for vessel services in the European market; a sister vessel of the one used at Sheringham Shoal received a three-year contract for work in the oil industry at a dayrate of $300,000/day. We expect that dayrates exceeding $300,000/day will be rare in the U.S. market and will only occur if vessel supply is severely constrained.

Greater Gabbard

Seajacks signed a $62 million contract to install turbines at the 504 MW Greater Gabbard wind farm. Greater Gabbard is composed of 140, 3.6 MW turbines and is 20 nm from port. The expected installation time is 14 months and the effective dayrate is $147,000/day. Using model default values yields a cost of $61.1 million and work duration of 13.9 months. The dayrate from the contract terms gives an estimated cost of $66.7 million. In this case, the model output and contract costs match closely.