Cost optimized dynamic design of offshore windfarm transformers with reliability and contingency considerations

Application of Dynamic Thermal Rating (DTR) for optimal utilization of transmission system components has been addressed in several studies. For Offshore Wind Power Plants (OWPPs), as a consequence of intermittent wind generation combined with favorable ambient conditions and strict design regulations, power transformers tend to be over-dimensioned.

Experience suggests that transformer lifetime exceeds OWPP life by tens of years, therefore DTR-based design of transformers can facilitate the optimization of OWPP export system. The recent breakthroughs in Thermo-Electric Equivalent (TEE) and aging models of transformers suggest that temporal development of Top Oil (TOT) and Hot Spot (HST) temperatures can be calculated efficiently, which when combined with monitoring of moisture and oxygen development can accurately assess the transformer health in real-time using Degree of Polymerization (DP).

However, the economic optimization of offshore windfarms substations using DTR and lifetime utilization-based transformer design during the OWPP planning phase has not been addressed in the literature due to the risks involved in balancing system reliability with windfarm profitability. In this paper, a novel strategy is proposed to utilize transformer DTR for cost optimization of offshore windfarm export system during the OWPP design phase.

Net Present Value (NPV) is used to assess capital cost impacts on transformer and Offshore Substation (OSS) design in year 0, and yearly revenue change over OWPP lifetime due to increased losses and possible energy curtailment as a result of smaller transformers. The unique methodology accounts for varying wind turbine availability due to maintenance and concurs with OWPP design requirements for OSS transformer contingency by generating statistical scenarios based on distribution functions and Markov Chain (MC) models respectively.

Moreover, the MC models are adapted for two fundamentally different design concepts for OWPP export system: n-1 contingent and non-contingent. The implementation of probabilistic Markov models to resolve stressful periods with increased transformer loading is distinctive to this paper and offers a unique perspective for DTR-based transformer size optimization.

Long-term site assessment data for OWPP including historical wind generation is used for transformer thermal assessment (TOT and HST) based on its load and ambient conditions. Furthermore, the reliability of design for varying construction and commissioning conditions is ensured by tracking DP-based transformer lifetime utilization over the operation period for three scenarios of oxygen and moisture development: conservative, high moisture and high oxygen.

The case study of a 1200 MW offshore windfarm with two 220 kV parallel export circuits and four 33/220 kV transformers in one OSS located 50 km off the east coast of UK has been used in this paper. 100 scenarios of wind turbine availability and transformer contingency are simulated and the optimal design is determined for the worst-case scenario.

Results indicate that optimal transformer rating can result in high thermal stress and minor energy curtailment during extreme contingency conditions for both the design concepts. However, the reduction in transformer size to align with OWPP operation life can reliably improve the windfarm business case significantly, particularly for the n-1 contingent design concept with conservative transformer moisture and oxygen conditions.