TODAY’S STUDY: HOW TO DESIGN A WIND FARM

Applications of Systems Engineering to the Research, Design, and Development of Wind Energy Systems
K. Dykes and R. Meadows w/F. Felker, P. Graf, M. Hand, M. Lunacek, J. Michalakes, P. Moriarty, W. Musial, and P. Veers, December 2011 (National Renewable Energy Laboratory)

Executive Summary

Over the past 30 years, wind energy has evolved from a small industry active in a few countries to a large international industry involving major players in the manufacturing, development, and utility sectors. Coinciding with the industry growth, significant innovation in the technology has resulted in larger sized turbines with lower associated costs of energy and more complex designs in all subsystems—from the rotor to the drivetrain to the electronics and control systems. However, as the deployment of the technology grows and its role within the electricity sector has become more prominent, so have the expectations of the technology in terms of performance, reliability, and cost. For the industry to continue to succeed and become a sustainable source of electricity, innovation in wind energy technology must continue to improve performance and lower the cost of energy while supporting seamless integration of wind energy into the electric grid without creating significant negative impacts on local communities and environments. At the same time, the nature of the issues associated with wind energy design and development are noticeably more complex than in the past due to a variety of factors such as, for example, large turbines sizes, offshore deployment or complex terrains. Looking toward the future, the industry would benefit from an integrated approach that simultaneously addresses turbine design, plant design and development, grid interaction and operation, and mitigation of adverse community and environmental impacts. These activities must be integrated in order to meet this diverse set of goals while recognizing trade-offs that exist between them.

click to enlarge

In order to address these challenges, National Renewable Energy Laboratory (NREL) has embarked on the Wind Energy Systems Engineering (WESE) initiative to evaluate how methods of systems engineering can be applied to the research, design, and development of wind energy systems. Systems engineering is a field within engineering that has a long history of application to complex technical systems such as aerospace. As such, the field holds much potential for addressing critical issues that face the wind industry today. This paper represents a first step for understanding this potential and lays out a conceptual design for the development of a WESE framework and tool. It reviews systems engineering methods as applied to related technical systems and illustrates how these methods can be combined in a WESE framework to meet the research, design, and development needs for the future of the industry. Subsequent efforts will focus on developing and implementing a framework based on the conceptual design illustrated in the last chapter of this report.

In general, systems engineering approaches have the following four characteristics: holistic, multidisciplinary, integrated/value-driven, and long-term/life-cycle oriented. The approach is holistic in that it considers the full technical system, including any number of performance criteria, as well as potentially non-technical concerns related to human factors or societal impacts. Systems engineering work is multidisciplinary, involving engineering, natural, computational, and even social sciences. It is also integrated and value-driven by considering the needs and interests of all customers and stakeholders.

click to enlarge

Finally, systems engineering is focused on the long-term or life cycle of the system and takes into account the cradle-to-grave life of the system. Beyond these four primary traits of the field, three common characteristics of the large-scale, complex technical systems are the focus of systems engineering work. These include complexity, uncertainty, and heterogeneity. The key characteristics of large-scale, complex technical systems also align with key attributes of wind energy systems, including the following:

• Complexity: Wind energy involves nearly every field of engineering and many of the natural and social sciences. The design of a wind turbine and plant interlinks these distinct disciplines for a holistic and multidisciplinary design that integrates the interests of a wide variety of stakeholders for operation over years and decades.

• Uncertainty: The science of wind and wind energy technology is still evolving. An incomplete understanding of both the physical processes and their interaction with the technology leads to an uncertain design environment. Even if a complete understanding of the system was obtained, there would still be uncertainty affecting system design, for example, with respect to the behavior of weather over time as it would impact a particular turbine or farm. Finally, there are external sources of uncertainty, such as political and economic developments, that can drastically affect the financial viability of wind energy projects.

• Heterogeneity: Wind turbines and plants must be designed for and operated in a wide array of environments—both from a physical standpoint and from an economic, social, and political standpoint. The U.S. Department of Energy (DOE) has separated factors that limit wind energy development into those that affect the cost of energy and those that impose market barriers such as social, environmental, and political factors.

The scope of wind energy design can be illustrated by a map of major design variables within a wind energy system as shown in Figure 1.

click to enlarge

The development and deployment of wind energy systems is affected by physical design drivers and impacts associated with various stakeholders from suppliers to original equipment manufacturers (OEM), developers, financiers, utilities, environments, and communities. A view of and an approach to wind energy research, design, and development must take all of these diverse factors into account. Design of these complex, uncertain, and heterogeneous large-scale technical systems is well suited to a systems engineering approach that is holistic, multidisciplinary, integrated, and life cycle oriented.

This paper addresses a wind energy system that falls geographically within a wind plant.

This includes all of the components, individual wind turbines, and the interactions between them as well as balance of station and operations and maintenance. In essence, the scope includes the traditional set of design drivers that are considered in looking at wind plant cost of energy. The methods reviewed in the paper reflect this scope, although the ultimate goal is to incorporate design objectives and methods that relate to the entire wind energy system, including grid interaction, community, and environmental impacts.

This paper surveys the landscape of systems engineering methods and catalogues the various existing modeling tools that relate to the design of wind energy systems from components to entire plants. It then provides an overview of how the existing set of design tools as well as future extensions may be coupled together within a systems engineering framework that will provide for a large variety of potential applications at the frontier of wind energy research, design and development. Examples of such applications that are relevant to future wind energy development may include:

• Optimization of the full wind plant system to achieve the lowest cost of energy and improve wind generated electricity costs relative to other generation technologies while maintaining or improving annual energy production. Use of systems engineering techniques such as multidisciplinary design optimization could yield plant designs that achieve significant system cost reduction by accumulating cost reduction across a number of components and consider long-term operation and maintenance impacts.

• Trade studies to evaluate different design concepts that are needed to prioritize R&D efforts to develop fixed-bottom and floating platform offshore wind plants. Use of systems engineering techniques such as multi-objective optimization and tradespace exploration could be valuable in assessing widely different concepts, including VAWTs.

click to enlarge

As offshore wind technology infrastructure is developed in the US, supply chain analysis could be used to evaluate various port facilities along with technology options for installation and servicing wind projects. Combined with optimization of technology designs to accommodate port facility requirements, projects with the lowest cost of energy would be identified.

Many wind energy specific design tools and methods already exist to address aspects of the illustrated challenges listed above, but a systems approach is needed to tackle them with adequate rigor. Integrating these tools into a SE framework that (1) permits comprehensive analysis using well developed SE methods and (2) is designed to expand systematically to create an ability to address the issues above and will aid the wind industry in achieving the next generation of lower cost wind technology.

click to enlarge

Methods in Systems Engineering

To address how a systems engineering approach might be used to design a wind turbine and/or plant, a survey was compiled of methods within systems engineering that may be applicable or that have already been tested for use in wind turbine and plant design. In general, these methods can be partitioned into three sets: design tools related to physical system design, methods related to supply chain and logistics, and other methods such as reliability and cost engineering. These methods will inform a WESE approach to the research, design, and development needs for the future of the industry.

Within the first set of tools related to physical system design, multidisciplinary design optimization (MDO) is featured because it has been used extensively in the design and research of aerospace and similarly complex technical systems. MDO allows the integration of different disciplinary design objectives into an overall system design optimization. It is a way to hierarchically decompose a complex design problem so that it maps better to existing partitions of disciplinary design efforts. MDO has been used already in a few research applications to wind energy that seek to optimize system cost of energy by integrating analysis across various disciplines including, for example, aerodynamics, structures and controls. In addition, the survey of methods for wind energy system design included a discussion of multi-objective optimization (MOO) that evaluates different design objectives either through a weighted technique or some sort of hierarchical ordering. Such methods are particularly appropriate when multiple stakeholders have conflicting interests for system design. MOO is also useful to evaluate a set of designs along various dimensions, such as wind energy system production, weights, reliability, etc. MOO techniques may result in a trade space of designs that can be compared using visual and statistical techniques. This may be used to evaluate trade-offs between different wind energy system designs or architectures rather than focusing on sensitivity to design parameters for a single detailed design.

In addition to turbine design, MDO methods can be extended to incorporate the entire wind plant and associated design objectives such as annual energy production, balance of station costs, and operations and maintenance costs. Looking at balance of station, supply chain considerations become important both in terms of initial plant design—including transportation, installation, and assembly logistics—but also to long-term O&M. Long-term development of a WESE framework may incorporate more advanced models for balance of station and O&M of wind plants that would integrate supply chain model techniques such as network analysis. With regard to plant operations and maintenance, reliability engineering and cost engineering are two methods that may also be applied to wind energy system design. Thus, a systems engineering approach to the research, design and development of wind energy systems may incorporate a variety of methods depending on system scope. The application of systems engineering methods at NREL will involve leveraging existing modeling tools within the development of a systems engineering specific tool.

click to enlarge

NREL Research and Design Tools for Wind Energy Systems

The integration and optimization of overall system properties within the wind system design toolset used at NREL is already a near-term goal at the NWTC. Figure 2 shows the current state of NREL wind energy system design tools as they relate to the systems engineering methods discussed in the previous section. The right side of Figure 2 shows the expected and desired development of the program and reflects the goal of integrating and developing existing tools within an overarching systems engineering framework. NREL is a primary developer of many different tools for the research, design and development of wind energy systems, including a suite of aeroelastic design codes for detailed time-series analysis of various turbine loads as well as a cost of energy model that estimates how design changes may affect everything from specific component costs to annual energy production to balance of station costs. In addition, there are various models developed at NREL related to external impacts of turbine and plant design including noise analysis tools as well as dynamic models of wind turbines for grid interaction. Many of these tools incorporate aspects of systems engineering from multidisciplinary analysis in the aeroelastic codes to supply chain analysis in the cost of energy and balance of station models. These characteristics have led to the use of such tools within several systems engineering research projects focused on wind energy systems. However, the integration of the tools within an explicit systems engineering toolset and framework will allow for a wide range of new and higher fidelity analyses that will improve the overall performance of wind energy systems.

click to enlarge

Development of a WESE Tool

The overall vision for a systems engineering design approach is to develop a framework and corresponding toolset that will allow for the integration of a variety of models for different aspects of the overall wind energy system. At all times, a WESE framework will maintain the capability for representing the full wind energy system including individual turbines and components, wind plants and turbine to turbine interaction, and cost of energy modeling for the plant and balance of station. Later realizations of a WESE framework may extend into advanced supply chain representations as well as grid integration and analytical capabilities for community and environmental impacts. The full system representation will have varying levels of model fidelity for each aspect of the system. Depending on the application, different models for each sub-system or discipline may be used in an overall analysis.

As higher fidelity models and improved design tools are developed, they will be integrated for use in a WESE framework. This will allow for continual evolution of the overall tool and increased fidelity of different system sub-models. For instance, a tool may initially incorporate a few different models of the turbine itself including aeroelastic design codes such as the Fatigue, Aerodynamics, Structures, and Turbulence (FAST) Code or the simplified WT_Perf or even parameterized metamodels. The tool might then be extended to interact with more detailed models for structural analysis of different turbine components that would interact with the full turbine model. Cost models might initially incorporate parameterized models such as the NREL model, other simplified models of turbine cost, or even an engineering-based cost model that is extended to capture detailed plant costs. Plant models might contain various levels of fidelity for modeling turbine interaction as well as site impact considerations on wind flows. The key aspect of development of a WESE tool is that it will allow for the integration of a range of models representing the different aspects of system design above and that these may be allowed to evolve over time. A high-level depiction of the models included in such a tool as well as the types of analyses to be performed is shown in the Figure 3.

click to enlarge

A project of this scope could easily become intractable if not managed in a systematic way. Therefore, within the large space of development, it is important to consider particular applications that would constitute a progression in tool development that is feasible in the near term. Specific steps of integration will build out a tool’s capabilities in terms of both model types and analytical methods. At the same time, a working toolset will be preserved at each step so that novel and useful analyses may be performed over the entire development of a tool. The potential steps to integration will reflect the current status of development of NREL wind energy analysis and design tools and also the needs associated with development of those tools. This will likely include four general phases:

1. Integration of physical turbine and cost of energy models for sensitivity analysis and optimization
2. Integration of detailed component design with physical turbine and cost of energy models for scenario analysis, trade studies and optimization

3. Integration of plant layout tools with the above set for full plant level analysis including for scenario analysis, trade studies and optimization
4. Integration of models for non-traditional design criteria such as utility, community, and environmental impacts.

click to enlarge

A WESE tool could support a diverse set of applications and a variety of analyses such as multidisciplinary optimization (with different optimization algorithms and techniques), multi-objective optimization, and development of trade spaces. This would allow for both detailed design optimization as well as evaluation across diverse system architectures. For example, a MDO may be used to perform a detailed cost optimization of a particular point design, and MOO may be used to survey a wide variety of different wind turbine configurations. In addition, the tool would include a range of post-processing decision-support tools including sensitivity analysis, uncertainty quantification, and visualization methods. User inputs would include input parameters for system design (turbine, plant and exogenous factors), but also would include the specific analyses to be performed and the models to be used (selection of sub-models based on level of fidelity and type of representation desired or the incorporation of user-defined sub-models). Every analysis thus would include: (1) a connection of individual elements and sub-systems into a system design space for a full system representation and (2) varying levels of fidelity in terms of modeling different subsystems depending on the chosen application. The overall development of such a tool will be a complicated process that involves the integration of disparate codes, obtaining metamodels of adequate fidelity, coupling across software packages, and other challenges. Thus, careful planning and management of the overall process upfront and at each step along the way will be important to the overall success of the initiative.

In summary, NREL’s wind energy systems engineering design initiative seeks to address a variety of issues that impact the current and future development of the wind energy sector. Wind electrical generation is a large-scale and complex technical system with various social impacts. As a result, a systems perspective and approach must be taken to the research, design, and development of these systems in order to meet the myriad of goals for future development of the technology. The inherent complexity of the physical system leads directly to a multidisciplinary approach to the design of the turbine itself, but also then to the plant level and beyond to the impacts that the plant will have on local utilities, communities, and environments. Systems engineering, which has a long history of development and application to a variety of industries, shows significant potential for addressing these system design challenges and will be a useful framework and tool for guiding and coordinating wind energy research, design and development activities among a variety of stakeholders including government, industry, national laboratories and academia.