In Data We Trust:
Rich Analytical Tools 
Optimize Project Life

Hybrid Energy Innovations 2015

On the gently sloping Montezuma Hills above Sacramento sits a striking symbol of the progress wind energy has made in the last 20 years.

A handful of 100 kW turbines atop diminutive lattice towers slice the air impatiently alongside the latest megawatt giants that have all but taken over this place.

In 2012, the replacement of 235 Kenetech 100 kW turbines with 50 REpower MM92-2.05 MW machines marked a new chapter for wind power in this corner of northern California. It was, at the time, the largest wind farm repowering ever completed in the U.S. It certainly will not be the last.

But just as some owners are repowering old wind farms with new turbine technology, others are extending the operating life of existing assets even further. When Portuguese utility company EDP Renewables performed a technical review of its fleet of 153 wind farms in 2011, a component lifetime fatigue analysis and risk review did not result in any wind farms being repowered. Rather, the study concluded that it was possible to extend the project life of those assets from 20 to 25 years, resulting in a $29 million positive impact on the company’s bottom line due to lower depreciation charges.

These two examples represent different approaches to the same challenge that is set to become increasingly important for a broad range of stakeholders in the North American wind business: maximizing financial returns and lowering the cost of energy (COE) for aging wind farms.

Fortunately, new analytical techniques and an increased use of operational data are greatly enhancing the ability of wind farm owners, operators and investors to identify optimized approaches to running older wind farms. The technical and economic impacts of decisions are becoming more data driven – and, therefore, more certain. This is an important step toward the continuing drive to lower the lifetime cost of wind energy.

A striking leap in generation on display in California’s Montezuma Hills.

 

Age-old question

The outward signs of progress in turbine design are starkly obvious in a place like the Montezuma Hills. Hub heights and blade lengths have grown radically, lattice towers replaced with smooth steel.

However, technological advancements inside turbines have led to the emergence of these more sophisticated lifecycle strategies. Advancements include the widespread adoption of SCADA systems capable of recording a broad range of operational parameters and, crucially, the development of techniques to extract practical value from these data. Parameters such as wind speed, turbulence intensity, generator speed and power, as well as drivetrain temperatures and vibration data, are being combined with systematic analysis tools capable of estimating accumulated fatigue to periodically assess a turbine’s remaining life.

Going one step further, today’s advanced turbine controllers are increasingly able to modify turbine operations based on the results of ongoing life assessment in order to maximize economic potential through the generating life of an asset. This ability to accurately understand the remaining useful life of wind turbines within a wind farm is central to developing better, more optimized lifecycle strategies. Full lifetime assessment enables the evaluation of different operation strategies and reveals the economic impact of each option.

At a basic level, this involves an analysis of the wind conditions experienced at a site compared to the environmental conditions for which those turbines were designed. A turbine whose design lifetime has been spent operating in benign conditions relative to its design parameters is likely to retain significant useful life despite an advanced age. Even turbines operating near each other at the same site can age at very different rates due to uneven exposure to factors such as turbulence, gusts and upflow. The design life of a turbine, therefore, does not necessarily coincide with the economic life, the point at which the smart business decision is to decommission and, potentially, repower with new technology.

More advanced life assessment involves mining huge amounts of SCADA data alongside a host of other information generated through the life of a wind farm in order to paint a more complete picture of turbine wear and the reasons for it. This can include condition monitoring data such as gearbox oil samples and acoustic sensor data, mechanical load measurements (where available), operations and maintenance (O&M) reports, or end-of-warranty inspections.

New sources of information, such as turbine-mounted LIDAR measurements, can add further sophistication to the analysis of how a turbine is interacting with its environment. Combined, these data sources can be used to quantify the relationships between fatigue loading of components and measured variables such as turbulence intensity, wind speed and air density. Cost models that link turbine component lifetimes to the actual loading experienced allow one to assess how site-specific factors, such as turbulence, directly impact economic life. Thus, more and better data mean greater certainty in the calculation of remaining economic life – with big implications for owners, operators and investors.

Data from wind turbines can optimize performance and maximize turbine life.

 

Optimization strategies

Armed with the best possible knowledge of the remaining useful life, detailed cost modeling can then be used to develop and evaluate a range of different lifecycle strategies. This provides a practical, data-driven framework to evaluate options such as turbine control modifications (hardware- or software-based), project control strategies, wind sector management, O&M approaches, up-rating, and inspection campaigns, as well as if and when to retrofit, decommission or repower.

For each strategy, the impacts on remaining useful life and on O&M costs are modeled using a probabilistic approach and Monte Carlo simulation to quantify uncertainties. The results can be evaluated using a time-series-based economic model that calculates the net present value (NPV) and COE resulting from each scenario and compared in order to assess which option provides the highest returns at acceptable risk.

The analysis also provides a platform for optimization. This approach enables those interested in understanding the implications and trade-offs among the full range of lifecycle options to design an operating strategy to meet specific investment goals.

Consider the following example: a probabilistic life assessment performed on a site with a large fleet of older, smaller turbines where the owner may be considering three scenarios – infilling with larger and newer turbines, modifying the control software to extend turbine life, and increasing O&M activities.

In this case, the life assessment indicated a high probability that infilling – when newer turbines are inserted between older turbines in operation – would reduce the remaining economic life of 16% of the smaller turbines from 20 years down to five years.

Control software modifications would extend the remaining project life to 22 years, and increasing O&M activities would extend the project life to 25 years. The probabilistic financial analysis shows that increasing O&M activities would result in the highest NPV for the project.

Whether the goal is short-term cashflow, project sale price, or maximized profits over 10 to 20 years due to tax appetites, production tax credits or power purchase agreements, lifecycle strategy optimization provides stakeholders with financial metrics (such as the NPV, internal rate of return or COE) that can be used to evaluate the costs and benefits of each scenario, given the specific constraints and objectives.

Some cases will be clear cut, such as when there is no remaining useful life and repowering is the best option. Other situations are more subtle, and the best approach will depend on finding a balance among competing factors.

 

Cost of energy

Increased collection and analysis of data from additional sensors and condition-based monitoring systems clearly comes at a cost, and for the smallest and oldest machines, the availability of more sophisticated data metrics will be limited.

Even then, with less data to work with, lifecycle optimization has been shown to deliver tangible economic returns. More importantly, because 90% of the operational wind capacity in the U.S. was installed within the last 10 years, the majority of currently aging assets are more sophisticated, megawatt-class turbines for which the economics of lifetime strategy planning are increasingly compelling.

Ultimately, the implementation and periodic review of lifecycle strategies allows owners to make smart decisions that maximize the value of their assets and deliver the lowest cost of energy to the end users. Furthermore, the greater certainty this type of strategic planning brings to investment decisions can help reduce financing costs.

By building a comprehensive understanding of a wind farm site through the synthesis of multiple data sources, the process of periodic lifecycle optimization also provides benefits that extend beyond the life of a single wind farm. These benefits include improved understanding of site wind flow characteristics, especially turbulence and wake impacts, and how these ultimately impact turbine wear. Another benefit is better knowledge of diurnal and seasonal production profiles. Even knowing the pattern of local weather and how this can impact O&M services provides valuable information for future developments. Combined, these factors allow greater certainty and lower associated development costs when the time to repower arrives.

In this sense, the small but pioneering turbines installed on the Montezuma Hills more than two decades ago are the real giants on whose shoulders a new generation of turbines gratefully stand. w

 

Craig Houston is a senior strategy and policy consultant and Ruth Heffernan Marsh is principal engineer at DNV GL. They can be reached at craig.houston@dnvgl.com and ruth.marsh@dnvgl.com, respectively.

Industry At Large: Lowering the Cost of Energy

In Data We Trust: Rich Analytical Tools Optimize Project Life

By Craig Houston & Ruth Heffernan Marsh

How data-driven decision-making is assisting owners and operators in lowering the cost of wind energy.

 

 

 

 

coverphoto.indd

NAW_body hyperlink NAW_body_i NAW_body_bi NAW_body_b_i NAW_body_b

NAW_first_graph

NAW_depbio

NAW_sub

NAW_last_graph

AuthorBio

NAW_SH

NAW_SH_no_rule

NAW_SH norule

NAW_SH_norule

NAW_SH_first_item

pullquote

sidebar_headline