Prevention of green energy loss: Estimation of fire hazard potential in wind turbines

Sun, W.; Lin, Wen-Yinn; You, Fei; Shu, Chi‐Min; Qin, Shenghui

doi:10.1016/j.renene.2019.03.045

Cited by 12 publications

(4 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Changes in the frequency and size of wildfires in the Mojave Desert are linked to the expansion of brome grasses [59]. Years with high invasive cover near wind turbines is a concern from a fire risk standpoint since turbines are a known source for fire ignition [17,60]. In the summer of 2012, for example, there were two separate instances of turbines igniting and causing grass fires, one in the Tehachapi site and one in the Desert Hot Springs site [61].…”

Section: Discussionmentioning

confidence: 99%

“…Once established, non-native annual grasses can create a continuous layer of fine fuels in otherwise sparsely covered desert plant communities and facilitate a cycle of fire in ecosystems not adapted to fire [16]. Energy development infrastructure and materials are also ignition sources [17], potentially helping to accelerate the grass-fire cycle in these areas. The direct relationship between land change driven by EED and the establishment and spread of invasive species in the arid Southwest has not yet been well quantified.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Landsat Time Series Assessment of Invasive Annual Grasses Following Energy Development

2019

View full text Add to dashboard Cite

Invasive annual grasses are of concern in much of the western United States because they tolerate resource variability and have high reproductive capacity, with propagules that are readily dispersed in disturbed areas like those created and maintained for energy development. Early season invasive grasses “green up” earlier than most native plants, producing a distinct pulse of greenness in the early spring that can be exploited to identify their location using multi-date imagery. To determine if invasive annual grasses increased around energy developments after the construction phase, we calculated an invasives index using Landsat TM and ETM+ imagery for a 34-year time period (1985–2018) and assessed trends for 1755 wind turbines installed between 1988 and 2013 in the southern California desert. The index uses the maximum Normalized Difference Vegetation Index (NDVI) for early season greenness (January-June), and mean NDVI (July–October) for the later dry season. We estimated the relative cover of invasive annuals each year at turbine locations and control sites and tested for changes before and after each turbine was installed. The time series was also mapped across the region and temporal trends were assessed relative to seasonal precipitation. The results showed an increase in early season invasives at turbine sites after installation, but also an increase in many of the surrounding control areas. Maps of the invasive index show a region-wide increase starting around 1998, and much of the increase occurred in areas surrounding wind development sites. These results suggest that invasions around the energy developments occurred within the context of a larger regional invasion, and while the development did not necessarily initiate the invasion, annual grasses were more prevalent around the developments.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Landsat Time Series Assessment of Invasive Annual Grasses Following Energy Development

2019

View full text Add to dashboard Cite

show abstract

“…GridSearchCV was chosen for this study due to its ability to systematically explore the hyperparameter space and identify the optimal combination of hyperparameters. Max_depth is searched in an interval of (5,15) with a step size of 2; the learning_rate is searched in an interval of (0.1, 1) with a step size of 0.1; and the search interval of n_estimators is searched in an interval of (1, 100) with a step size of 10. These ranges and step sizes were chosen based on the authors' experience and initial experiments to balance computational efficiency and thoroughly explore the hyperparameter space.…”

Section: Model Tuning Methodsmentioning

confidence: 99%

“…This method can warn of fires caused by overcurrent, but it has difficulty warning of other types. Sun et al [15] used Fourier transform infrared spectroscopy to analyze the effect of functional group changes in hydraulic oil, gear oil, and transformer oil on temperature during wind turbine operation to provide early warnings of fire. However, an entire wind turbine has not been studied.…”

Section: Introductionmentioning

confidence: 99%

Using Transfer Learning and XGBoost for Early Detection of Fires in Offshore Wind Turbine Units

Wan,

Du,

Gong

et al. 2024

Energies

View full text Add to dashboard Cite

To improve the power generation efficiency of offshore wind turbines and address the problem of high fire monitoring and warning costs, we propose a data-driven fire warning method based on transfer learning for wind turbines in this paper. This paper processes wind turbine operation data in a SCADA system. It uses an extreme gradient-boosting tree (XGBoost) algorithm to build an offshore wind turbine unit fire warning model with a multiparameter prediction function. This paper selects some parameters from the dataset as input variables for the model, with average cabin temperature, average outdoor temperature, average cabin humidity, and average atmospheric humidity as output variables. This paper analyzes the distribution information of input and output variables and their correlation, analyzes the predicted difference, and then provides an early warning for wind turbine fires. This paper uses this fire warning model to transfer learning to different models of offshore wind turbines in the same wind farm to achieve fire warning. The experimental results show that the prediction performance of the multiparameter is accurate, with an average MAPE of 0.016 and an average RMSE of 0.795. It is better than the average MAPE (0.051) and the average RMSE (2.020) of the prediction performance of a backpropagation (BP) neural network, as well as the average MAPE (0.030) and the average RMSE (1.301) of the prediction performance of random forest. The transfer learning model has good prediction performance, with an average MAPE of 0.022 and an average RMSE of 1.469.

show abstract

Thermal and non-thermal fire hazard characteristics of wind turbine blades

Wang,

Cheng,

You

et al. 2024

J Therm Anal Calorim

View full text Add to dashboard Cite

Prevention of green energy loss: Estimation of fire hazard potential in wind turbines

Cited by 12 publications

References 35 publications

Landsat Time Series Assessment of Invasive Annual Grasses Following Energy Development

Landsat Time Series Assessment of Invasive Annual Grasses Following Energy Development

Using Transfer Learning and XGBoost for Early Detection of Fires in Offshore Wind Turbine Units

Thermal and non-thermal fire hazard characteristics of wind turbine blades

Contact Info

Product

Resources

About