AT1 DSI Wind Energy
1. Introduction:
Wind Energy is the process of using air flow though wind turbines to create mechanical power. This mechanical power is then used to turn generators which create electricity. Wind is a naturally replenishing resource that occurs due to the uneven heating of the earth’s atmosphere (Fthenakis, V, Kim, H. C. 2009). As companies and individuals become more conscious of issues like global warming and the effect of their own carbon footprint, renewable energy from sources like wind have moved to the forefront of social awareness and public policy. Because this changing shift in public perception technology companies like Google and Apple are increasing their investments to achieve 100 percent power for their data centers from renewable energy (Etherington, 2017).
As wind energy becomes more economically competitive, wind farm operators must understand and manage the performance analysis of their farms in order to achieve desired production and revenue goals. There are a number of challenges currently facing the Wind Energy industry including the relative cost required to generate electricity in comparison to traditional fossil fuels, the day to day variability of output due to the erratic nature of wind and understanding and analysing the vast amounts of data created from sensors connected to wind turbines complex systems.
With public perception of energy production changing and wind energy aiming to establish itself as a viable alternative to fossil fuel based energy, a significant focus is being placed on the need to increase output and reduce cost. The use of big data and analytics is central to this process where data driven predictive models are being used. Increased performance and reduced cost is being made possible by the implementation of SMART wind farms and High-Fidelity modelling (Dykes, 2017).
2. Opportunities and Challenges facing Wind Energy
Cost
The higher cost of producing wind energy relative to fossil fuel based alternatives like coal represents a significant challenge for the industry as it tries to gain greater market share in the current “energy mix” in Australia. In 2017 the cost of producing wind energy in Australia was $60 per MWh as opposed to existing black coal-fired stations at $40 per MWh (Department of Environment and Energy, Australian Energy Update 2018). in 2016–17 wind energy accounted for 4.9% of the total electricity generated in Australia while coal was the largest source of energy generation in the country contributing approximately 62.7% (Department of Environment and Energy, Australian Energy Update 2018). It’s difficult for the Wind Energy industry to take a greater proportion of market share of the electricity production mix when such a large gap exists between the costs of the current methods of electricity production.
Coal-fired power stations in Australia have operating lives of around fifty years. Presently, nine of Australia’s twelve largest operating coal-fired power stations are more than 30 years old. In preparation for the retirement of these older coal-fired power stations, policymakers and energy companies are debating what types of electricity generation facilities will be commissioned to replace them (Department of Environment and Energy, Australian Energy Update 2018). Reducing the production costs of wind energy will enable the industry to take advantage of Australia’s electricity generation void that will present itself over the next two decades.
Currently, engineers are continuing to make progress in wind turbine design, but they have reached a point of diminishing returns. Significant research and development costs are needed to achieve small increases in turbine output. So at the moment the industry is looking towards other alternatives to improve efficiency (Dykes, 2017).
Variable nature of wind
A central issue the Wind Energy industry has continued to grapple with is the nature of wind and the fact that it behaves erratically. On any given day changes in atmospheric pressure may create powerful gusts of wind or it may create none at all. Unlike electricity from traditional fossil fuel based sources, the power flow from wind generators is erratic as winds rise and fall. This leads to an uneven output from wind farms and makes it tougher for energy producers to work seamlessly with power grids that must send a steady flow of electricity to companies and households (Winslow, 2017).
Wind is also not abundant and evenly distributed across the country. Wind Energy is generally generated away from urban areas because wind patterns are more difficult to measure and buildings obstruct and deflect wind, leading to increased turbidity and decreased intensity of the wind. Wind Energy creation works best in environments with strong and consistent winds, such as large open plains. Because the ideal environments for creating wind energy exist away from urban populations, there are additional costs associated with storing and transferring this electricity to urban areas where it is consumed (Winslow, 2017).
Analysing the vast amounts of data created from sensors
With the arrival of embedded sensors and advanced analytics in the Wind Energy industry the data generated from operations and maintenance produces an estimated 25 trillion bytes of information daily (Bouqata, 2017). This enormous amount of available data has created a shift from an information poor to an information rich industry. The embedded sensors in the wind turbines contribute to the plethora of data available. Embedded sensors accurately predict and solve impending failures increasing the lifespan of the turbines. Most importantly, the sensors provide accurate data about every turbine and its most important components which can be linked together and connected to a command center.
The challenge with the volume of data being created is to harness the useful data so it can be used in the development of an effective predictive model. Continuously collecting data from wind turbines combined with other operational data can allow wind farm operators to better understand what is happening in the field, plan ahead, and predict operating life, resulting in reduced maintenance costs and improved performance (Bouqata, 2017).
3. Opportunities for the use of Big Data to drive innovation
Big data analytics techniques can significantly improve wind farm performance and reduce costs. The use of advanced analytics for knowledge discovery, particularly machine learning, has emerged as a means to enable smart decisions (Bouqata, 2017). It has been used successfully to address problems in various industries, resulting in disruptive innovation and can be leveraged in the wind energy industry to solve challenges related to performance and maintenance costs.
One of the ways that engineers can use data to drive innovation in the Wind Energy industry is through the implementation of wind farms based on the SMART (System Management of Atmospheric Resource through Technology) principle. The turbines in SMART wind farms are designed to communicate and cooperate with each other which enables them to work as a team rather than a collection of individual turbines. At the moment, wind farms are made up of large turbines spaced far apart, with each turbine facing directly into the wind allowing it to absorb as much wind as possible. As more and more data has been collected on these farms, researchers have discovered that because the turbines at the front of the wind farm grab as much wind as possible, the entire operation produces less electricity overall (Dykes, 2017).
A SMART wind farm will respond to real time weather conditions, making individual turbines make small adjustments to the wind and adjust their blade pitch in order to steer the turbulent wake winds away from downstream turbines, allowing each turbine to perform at their best. The new SMART wind farms will incorporate turbines of varying sizes and heights, to optimise production for each unique location (Bouqata, 2017).
Another technique that can significantly improve wind farm performance and reduce costs is High-Fidelity Modeling (HFM) which is made possible by advances in sensors and computer technology. High-fidelity modelling uses a set of transducers to measure a number of parameters including wind direction, speed, and turbulence, as well as the condition of the turbines themselves. Using data obtained from these sensors, wind farm controllers can maximise electricity production by directing wind flow based on current conditions. By monitoring each turbine’s operation, high-fidelity modeling can predict component failures, measure turbine loads, and optimise maintenance schedules, which increases the efficiency and output and lowers the overall cost of operating a wind farm (Dykes, 2017).
High-fidelity modeling helps scientists and researchers more accurately measure, describe, and predict wind patterns. Equipped with this scientific data, engineers can design wind farms so that renewable energy can provide consistent electricity generation to the grid and factor in irregularities like peak demand response, power factor correction, and frequency regulation (Dykes, 2017).
4. Potential data issues
The ability of the Wind Energy industry to collect quality data is essential to the development of an effective predictive model. Developing an effective predictive model to increase efficiency and reduce cost is extremely complex due to the model being impacted by two factors of wind scale. The inflow of wind into the farm which is ultimately driven by weather data as well as the subsequent flows of wind within the plant which is affected by the collective behaviour of each of the wind turbines (Dykes, 2017).
The complexity of accurately predicting these two scales of wind imposes limitations on current wind turbine design and innovation because actual conditions often differ from what can be predicted. These uncertainties impact the ability of the wind industry to develop more integrated and optimised approaches to wind power plant design and operation. The multiple atmospheric and physical processes involved make the wind modeling problem a difficult challenge. Although researchers have made great improvements in understanding the flow of wind at each scale, there is still great ambiguity in the physics of each scale as well as the interactions of flow between the two. The interactions between the two scales requires advance analytics and efficient machine learning algorithms to resolve (Dykes, 2017).
5. Impact on the Wind Energy Sector
Leveraging big data can enable the Wind Energy industry to more accurately measure, describe, and predict wind patterns. With this scientific data, engineers can design wind farms so that renewable energy can provide more consistent, reliable electricity generation at a cheaper price (Bouqata, 2017).
Through better analysis of these large volumes of data that is becoming available the implementation and design of SMART Wind farms allows operators to better understand what is happening in the field so they can plan ahead ultimately resulting in improved performance and reduced maintenance costs.
Bibliography:
Andrew R. Winslow
2017 Urban Wind Generation: Comparing Horizontal and Vertical Axis Wind Turbines at Clark
(Bouchra Bouqata, 2017) Big Data and Analytics for Wind Energy Operations and Maintenance: Opportunities, Trends, and Challenges in the Industrial Internet
https://www.ncbi.nlm.nih.gov/books/NBK481626/
Fthenakis, V.; Kim, H. C. (2009). “Land use and electricity generation: A life-cycle analysis”. Renewable and Sustainable Energy Reviews.
Etherington D., 2016 Google says it will hit 100% renewable energy by 2017
https://techcrunch.com/2016/12/06/google-says-it-will-hit-100-renewable-energy-by-2017/
Katherine Dykes 2017, Enabling the SMART Wind Power Plant of the Future Through Science-Based Innovation,. National Renewable Energy Laboratory
Department of the Environment and Energy (2018) Australian Energy Update, Canberra, August.
https://www.energy.gov.au/sites/default/files/australian_energy_update_2018.pdf
