Detection of Machinery Failure Signs From Big Time-Series Data Obtained by Flow Simulation of Intermediate-Pressure Steam Turbines

Abstract The periodic maintenance, repair, and overhaul (MRO) of turbine blades in thermal power plants are essential to maintain a stable power supply. During MRO, older and less-efficient power plants are put into operation, which results in wastage of additional fuels. Such a situation forces thermal power plants to work under off-design conditions. Moreover, such an operation accelerates blade deterioration, which may lead to sudden failure. Therefore, a method for avoiding unexpected failures needs to be developed. To detect the signs of machinery failures, the analysis of time-series data is required. However, data for various blade conditions must be collected from actual operating steam turbines. Further, obtaining abnormal or failure data is difficult. Thus, this paper proposes a classification approach to analyze big time-series data alternatively collected from numerical results. The time-series data from various normal and abnormal cases of actual intermediate-pressure steam-turbine operation were obtained through numerical simulation. Thereafter, useful features were extracted and classified using K-means clustering to judge whether the turbine is operating normally or abnormally. The experimental results indicate that the status of the blade can be appropriately classified. By checking data from real turbine blades using our classification results, the status of these blades can be estimated. Thus, this approach can help decide on the appropriate timing for MRO.

Download Full-text

Comparative Analysis of Low-Grade Heat Utilization Methods for Thermal Power Plants with Back-Pressure Steam Turbines

Energies ◽

10.3390/en14248519 ◽

2021 ◽

Vol 14 (24) ◽

pp. 8519

Author(s):

Nikolay Rogalev ◽

Vladimir Kindra ◽

Ivan Komarov ◽

Sergey Osipov ◽

Olga Zlyvko ◽

...

Keyword(s):

Ambient Temperature ◽

Boiling Point ◽

Power Plants ◽

Thermal Power ◽

Back Pressure ◽

Low Grade ◽

Thermal Power Plants ◽

Steam Turbines ◽

Pressure Steam ◽

Steam Production

Thermal power plants (TPPs) with back-pressure steam turbines (BPSTs) were widely used for electricity and steam production in the Union of Soviet Socialist Republics (USSR) due to their high efficiency. The collapse of the USSR in 1991 led to a decrease in industrial production, as a result of which, steam production in Russia was reduced and BPSTs were left without load. To resume the operation of TPPs with BPSTs, it is necessary to modernize the existing power units. This paper presents the results of the thermodynamic analysis of different methods of modernization of TPPs with BPSTs: the superstructure of the steam low-pressure turbine (LPT) and the superstructure of the power unit operating on low-boiling-point fluid. The influence of ambient temperature on the developed cycles’ efficiency was evaluated. It was found that the usage of low-boiling-point fluid is thermodynamically efficient for an ambient temperature lower than 7 °C. Moreover, recommendations for the choice of reconstruction method were formulated based on technical assessments.

Download Full-text

Influence of the reheat temperature on the efficiency of the wet-steam turbines of thermal power plants that operate under variable conditions

Thermal Engineering ◽

10.1134/s0040601517030077 ◽

2017 ◽

Vol 64 (4) ◽

pp. 295-300

Author(s):

A. A. Tarelin

Keyword(s):

Power Plants ◽

Thermal Power ◽

Thermal Power Plants ◽

Steam Turbines ◽

Wet Steam

Download Full-text

Steam turbines of the T-50/60-8.8, K-63-8.8, and Tp-100/110-8.8 types destined for modernization of thermal power plants with K-50-90 and K-100-90 turbines

Thermal Engineering ◽

10.1134/s0040601512120129 ◽

2012 ◽

Vol 59 (12) ◽

pp. 907-912 ◽

Cited By ~ 1

Author(s):

A. Ye. Valamin ◽

A. Yu. Kultyshev ◽

Yu. A. Sakhnin ◽

M. V. Shekhter ◽

M. Yu. Stepanov

Keyword(s):

Power Plants ◽

Thermal Power ◽

Thermal Power Plants ◽

Steam Turbines

Download Full-text

Explainable Deep Neural Networks for Multivariate Time Series Predictions

Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence ◽

10.24963/ijcai.2019/932 ◽

2019 ◽

Cited By ~ 8

Author(s):

Roy Assaf ◽

Anika Schumann

Keyword(s):

Neural Networks ◽

Time Series ◽

Network Architecture ◽

Power Plants ◽

Deep Neural Networks ◽

Time Series Data ◽

Multivariate Time Series ◽

Average Energy ◽

Series Data ◽

Time Interval

We demonstrate that CNN deep neural networks can not only be used for making predictions based on multivariate time series data, but also for explaining these predictions. This is important for a number of applications where predictions are the basis for decisions and actions. Hence, confidence in the prediction result is crucial. We design a two stage convolutional neural network architecture which uses particular kernel sizes. This allows us to utilise gradient based techniques for generating saliency maps for both the time dimension and the features. These are then used for explaining which features during which time interval are responsible for a given prediction, as well as explaining during which time intervals was the joint contribution of all features most important for that prediction. We demonstrate our approach for predicting the average energy production of photovoltaic power plants and for explaining these predictions.

Download Full-text

The Influence of Terminology on the Effective Solution of the Coal Ash Handling Problem

KnE Materials Science ◽

10.18502/kms.v2i2.959 ◽

2017 ◽

Vol 2 (2) ◽

pp. 133

Author(s):

V.Y. Putilov ◽

I.V. Putilova ◽

H-J Feuerborn

Keyword(s):

Power Plants ◽

Fossil Fuel ◽

Thermal Power ◽

Coal Ash ◽

Thermal Power Plants ◽

Effective Solution ◽

Global Trends ◽

Best Available Technologies ◽

The Status ◽

Definition Of

<p class="TTPTitle"><span>The paper presents the data on a structure of conventional fuels consumption, as well as a structure of fossil fuel consumption at thermal power plants (TPPs) in Russia. The issue of applying the Best Available Technologies (BAT) in Russia is touched upon. Statistics on production and utilization of coal ash in Russia and other countries is given. The paper provides information on the status and terminology of coal ash in different countries. It contains an impact of terminology on the effectiveness of solving the coal ash handling problem in Russia. The paper includes a new legislative definition of coal ash produced at thermal power plants and boiler houses of Russia which meets modern conditions and global trends, as well as requirements for coal energy sector.</span></p>

Download Full-text

ANALYZING THE STATUS OF MANGO TREES IN BRGY. CANTIPAY, CARMEN, CEBU USING NDVI AND TIME SERIES CLUSTERING

ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences ◽

10.5194/isprs-archives-xlii-4-w19-313-2019 ◽

2019 ◽

Vol XLII-4/W19 ◽

pp. 313-317

Author(s):

R. B. Navaja ◽

F. P. Campomanes ◽

C. L. Patiño ◽

M. J. L. Flores

Keyword(s):

Time Series ◽

Hierarchical Clustering ◽

Vegetation Index ◽

Time Series Data ◽

Minimum Variance ◽

Series Data ◽

Linkage Algorithm ◽

Division 1 ◽

The Status ◽

Cluster 2

Abstract. The Department of Agriculture – Region VII reports that many mango orchards in Cebu province are dying because of the absence of required post-harvest attention. Lacklustre yields and erratic pest infestations have driven some farmers and growers to abandon mango orchards. To help revive low-yielding mango orchards, there is a need to distinguish actively bearing mango trees from those that remain dormant throughout the year. Using remote sensing techniques, mango trees from separate orchards in Brgy. Cantipay, Carmen, Cebu were mapped and studied using multi-temporal Sentinel-2 data (from January 2018 through May 2019). Prior to that, a field visit was conducted to survey the area using UAVs and field observation, and in the process, was able to identify an abandoned mango orchard. Pixel-based Normal Difference Vegetation Index (NDVI) values were extracted from each of the 822 geotagged mango trees with an average of 16 trees among 53 divisions. Time series were derived from the average of the NDVI values from each division and plotted per month of extraction from oldest to latest. Clustering was applied to the time series data using Hierarchical Clustering with Ward’s Minimum Variance as an algorithm to determine the divisions with the closest time series. Using the resulting dendrogram as basis, two major clusters were selected based on the value of their distances with each other: Cluster 1 containing 29 Divisions, and Cluster 2 containing 24 Divisions. Cluster 1 contains most of the Divisions in and around the biggest active mango orchard. In contrast, Cluster 2 contains most of the Divisions that are in and around the previously identified abandoned mango orchard. An alternative dendrogram was also created by using Complete Linkage algorithm in Hierarchical Clustering, after which 3 relevant clusters were selected. The second dendrogram highlights the stark difference between Division 1, contained in Cluster 3, from the rest of the other clustered divisions at 2.17 units from the next closest one. Notably, Division 1 is located smack in the middle of the abandoned orchard The remaining clusters, Cluster 2 with 21 divisions containing most of the divisions in the abandoned orchard, is 2.46 distance units away from Cluster 1, which has 31 and hosting most of the divisions in the active mango orchards. Two major clusters emerged from using the two algorithms. Divisions with higher and more variant NDVI values seemed to come from the mango trees which were more active during the fruiting cycle. Divisions from the abandoned mango orchards were observed to have lower and less varied NDVI values because of minimal activity in the trees. Other Divisions clustered under the abandoned orchard could have been juveniles based on their size.

Download Full-text

Population Cycles of the Larch Budmoth in Switzerland

Population Cycles ◽

10.1093/oso/9780195140989.003.0011 ◽

2002 ◽

Author(s):

Peter Turchin ◽

Cheryl J. Briggs

Keyword(s):

Time Series ◽

Food Quality ◽

Time Series Data ◽

Population Cycles ◽

Series Data ◽

Food Quantity ◽

The Status ◽

The Rich ◽

Rich Data

The population dynamics of the larch budmoth (LBM), Zeiraphera diniana, in the Swiss Alps are perhaps the best example of periodic oscillations in ecology (figure 7.1). These oscillations are characterized by a remarkably regular periodicity, and by an enormous range of densities experienced during a typical cycle (about 100,000-fold difference between peak and trough numbers). Furthermore, nonlinear time series analysis of LBM data (e.g., Turchin 1990, Turchin and Taylor 1992) indicates that LBM oscillations are definitely generated by a second-order dynamical process (in other words, there is a strong delayed density dependence—see also chapter 1). Analysis of time series data on LBM dynamics from five valleys in the Alps suggests that around 90% of variance in Rt is explained by the phenomenological time series model employing lagged LBM densities, R, =f(Ni-1,Ni-2,) (Turchin 2002). As discussed in the influential review by Baltensweiler and Fischlin (1988) about a decade ago, ecological theory suggests a number of candidate mechanisms that can produce the type of dynamics observed in the LBM (see also chapter 1). Baltensweiler and Fischlin concluded that changes in food quality induced by previous budmoth feeding was the most plausible explanation for the population cycles. During the last decade, the issue of larch budmoth oscillations was periodically revisited by various population ecologists looking for general insights about insect population cycles (e.g., Royama 1977, Bowers et al. 1993, Ginzburg and Taneyhill 1994, Den Boer and Reddingius 1996, Hunter and Dwyer 1998, Berryman 1999). These authors generally concurred with the view that budmoth cycles are driven by the interaction with food quality. A recent reanalysis of the rich data set on budmoth population ecology collected by Swiss researchers over a period of several decades, however, suggested that the role of parasitism is underappreciated (Turchin et al. 2002). Before focusing on the roles of food quality and parasitism in LBM dynamics, we briefly review the status of other hypotheses that were discussed in the literature on LBM cycles. First, the natural history of the LBM-larch system is such that food quantity is an unlikely factor to explain LBM oscillations.

Download Full-text