Prediction Of Power Production Research Articles

Maximization of the annual energy production of wind power plants by optimization of layout and yaw-based wake control

This paper presents a wind plant modeling and optimization tool that enables the maximization of wind plant annual energy production (AEP) using yaw-based wake steering control and layout changes. The tool is an extension of a wake engineering model describing the steady-state effects of yaw on wake velocity profiles and power productions of wind turbines in a wind plant. To make predictions of a wind plant's AEP, necessary extensions of the original wake model include coupling it with a detailed rotor model and a control policy for turbine blade pitch and rotor speed. This enables the prediction of power production with wake effects throughout a range of wind speeds. We use the tool to perform an example optimization study on a wind plant based on the Princess Amalia Wind Park. In this case study, combined optimization of layout and wake steering control increases AEP by 5%. The power gains from wake steering control are highest for region 1.5 inflow wind speeds, and they continue to be present to some extent for the above-rated inflow wind speeds. The results show that layout optimization and wake steering are complementary because significant AEP improvements can be achieved with wake steering in a wind plant layout that is already optimized to reduce wake losses. Copyright © 2016 John Wiley & Sons, Ltd.

Assessment of Wind Atlas Analysis and Application Program and computational fluid dynamics estimates for power production on a Jeju Island wind farm

The accuracy of power production predictions made using a linear model, Wind Atlas Analysis and Application Program, and the computational fluid dynamics model, Meteodyn WT, was assessed for a Seongsan wind farm on Jeju Island, South Korea. The actual data on power production from 10 Vestas V80 2 MW wind turbines were collected through a supervisory control and data acquisition system. Wind data were collected over a period of 1 year from a 70-m-height met mast which was about 1.6 km away from the wind farm. The two applications were used for estimating the annual energy production for both an entire wind farm and an individual wind turbine and accomplished this by assessing five different wind farm sizes ranging from 5 km × 5 km to 20 km × 20 km. The results showed that Meteodyn WT software performed better than the Wind Atlas Analysis and Application Program in predicting power production for all the five domain sizes studied. For the entire wind farm, the relative error in annual energy production prediction was from 2.5% to 3.5% when running Meteodyn WT, while ranging from 3.8% to 11.5% when running WindPRO. While a little overestimation in the capacity factor occurred using the WindPRO application, a little underestimation was found using Meteodyn WT.

Comparison of Photovoltaic plant power production prediction methods using a large measured dataset

Nowadays the estimation of power production yield by stand-alone and grid-connected Photovoltaic (PV) plants is crucial for technical and economic feasibility design analyses. The main goal is to overcome renewables unpredictability by properly estimating the power production and by suitably balancing generation and consumption. In this context, many methods can be applied to forecast renewables energy production. The scope of this paper is a comparative analysis of three different methods to estimate the power production of a preexisting PV plant. It is installed at ENEA Research Centre located in Portici (South Italy) and it is integrated in a Micro Grid (MG) configuration. In detail a phenomenological model proposed by Sandia National Laboratories and two statistical learning models, a Multi-Layer Perceptron (MLP) Neural Network and a Regression approach, are compared. These models are deeply different also in terms of required input data and parameters. In detail, phenomenological model application requires the availability of design parameters and technical devices specifications. Statistical machine learning models need, however, input variable previously acquired datasets. The a-Si/μc-Si PV plant, installed at Portici, represents an adequate case study for the three models comparison, as both design and acquired data are available. In fact, the plant was designed at the ENEA Research Centre so this makes possible the knowledge of the design parameters and, being a part of the MG, its data are continuously acquired and transmitted to other network devices. Obtained results demonstrate more accurate power predictions can be reached by statistical machine learning approaches. The main novelty of the paper consists in the optimization of the considered models by the appropriate identification of the minimum and more representative training dataset. Authors underline the unnecessary use of thousands samples by suitably selecting the dataset size and samples by means of a Genetic Algorithm. The optimization strategy effectiveness is verified comparing the prediction performances obtained employing the optimal dataset with those obtained with a randomly chosen dataset. In this scenario, Genetic Algorithm strategy represents a successful approach to the suitable identification of statistical models datasets.

The Wind Forecast Improvement Project (WFIP): A Public–Private Partnership Addressing Wind Energy Forecast Needs

Abstract The Wind Forecast Improvement Project (WFIP) is a public–private research program, the goal of which is to improve the accuracy of short-term (0–6 h) wind power forecasts for the wind energy industry. WFIP was sponsored by the U.S. Department of Energy (DOE), with partners that included the National Oceanic and Atmospheric Administration (NOAA), private forecasting companies (WindLogics and AWS Truepower), DOE national laboratories, grid operators, and universities. WFIP employed two avenues for improving wind power forecasts: first, through the collection of special observations to be assimilated into forecast models and, second, by upgrading NWP forecast models and ensembles. The new observations were collected during concurrent year-long field campaigns in two high wind energy resource areas of the United States (the upper Great Plains and Texas) and included 12 wind profiling radars, 12 sodars, several lidars and surface flux stations, 184 instrumented tall towers, and over 400 nacelle anemometers. Results demonstrate that a substantial reduction (12%–5% for forecast hours 1–12) in power RMSE was achieved from the combination of improved numerical weather prediction models and assimilation of new observations, equivalent to the previous decade’s worth of improvements found for low-level winds in NOAA/National Weather Service (NWS) operational weather forecast models. Data-denial experiments run over select periods of time demonstrate that up to a 6% improvement came from the new observations. Ensemble forecasts developed by the private sector partners also produced significant improvements in power production and ramp prediction. Based on the success of WFIP, DOE is planning follow-on field programs.

Small-Scale Wind Turbine Testing in Wind Tunnels Under Low Reynolds Number Conditions

Much of the aerodynamic design of wind turbines is accomplished using computational tools such as XFOIL. These codes are not robust enough for predicting performance under the low Reynolds numbers found with small-scale wind turbines. Wind tunnels can experimentally test wind turbine airfoils to determine lift and drag data over typical operating Reynolds numbers. They can also test complete small wind turbine systems to determine overall performance. For small-scale wind turbines, quality experimental airfoil data at the appropriate Reynolds numbers are necessary for accurate design and prediction of power production.

An empirical approach to parameterizing photovoltaic plants for power forecasting and simulation

The aim of this work is to develop an algorithm that can utilize historical PV power measurements to establish the parameters of a physical model for power production. The chosen approach consists in evaluating the parameters of a PV model that maximize the likelihood that simulations match with power measurements. The proposed method offers advantages beyond the standard approaches used for the simulation or prediction of PV power production, as it makes maximum use of the information typically available on a PV plant (plant description and measurement history). Furthermore, an interpretation and control of the algorithm output is made possible. The performance of the proposed approach has been evaluated and analyzed using measurements from two PV plants. It is shown that the proposed approach may identify the orientation angles of a PV module to within an accuracy of less than 2° in optimal cases. Situations were also found with a difference between the estimated and actual angles of 5°, for which the estimated parameters lead to better simulation/forecast accuracy than the actual ones as they balance the systematic error of the chosen PV-model.

LES and RANS simulation of onshore Bessaker wind farm: analysing terrain and wake effects on wind farm performance

This work compares the predictive performance of RANS and LES solver in capturing the effect of terrain and wakes on the performance of the Bessaker wind farm. This 25 turbine wind farm is located in a highly complex terrain and is exposed to predominantly high westerly and south easterly winds. A one-equation sub-grid scale LES turbulence model has been used to help capture the wake dynamics: particularly the effects of wake meandering and wake-turbine interactions. A comparison between RANS and LES models highlights the influence of turbulence model on wake decay and its subsequent effect on prediction of power production. The LES model predicts delayed decay of the wake and more pronounced wake interference leading to a lower power production in wind farm than the RANS case. The RANS model overpredicts turbulence, which cause faster turbulent momentum diffusivity and faster wake recovery. This study has given some insights regarding the power production at Bessaker wind farm for neutral conditions and westerly flow.

Validation of the standalone implementation of the dynamic wake meandering model for power production

AbstractThis paper presents validation for using the standalone implementation of the dynamic wake meandering (DWM) model to conduct numerical simulations of power production of rows of wind turbines.The standalone DWM model is an alternative formulation of the conventional DWM model that does not require information exchange with an aeroelastic code. As a consequence, the standalone DWM model has significantly shorter computational times and lower demands on the user environment. The drawback of the standalone DWM model is that it does not have the capability to predict turbine loads. Instead, it should be seen as an alternative for simulating the power production of a wind farm. The main advantage of the standalone DWM model is the ability to capture the key physics for wake dynamics such as the turbine specific induction, the build‐up of wake turbulence and wake deficit in the wind farm, and the effect of ambient turbulence intensity and atmospheric stability.The predicted power production of the standalone DWM model is compared with that of full scale measurements from Horns Rev, Lillgrund, Nysted and Weingermeer wind farms. Overall, the difference between the models predictions and the reference data is on the order of 5%. Copyright © 2014 John Wiley & Sons, Ltd.

Photovoltaic power forecasting using statistical methods: impact of weather data

An important issue for the growth and management of grid-connected photovoltaic (PV) systems is the possibility to forecast the power output over different horizons. In this work, statistical methods based on multiregression analysis and the Elmann artificial neural network (ANN) have been developed in order to predict power production of a 960 kWP grid-connected PV plant installed in Italy. Different combinations of the time series of produced PV power and measured meteorological variables were used as inputs of the ANN. Several statistical error measures are evaluated to estimate the accuracy of the forecasting methods. A decomposition of the standard deviation error has been carried out to identify the amplitude and phase error. The skewness and kurtosis parameters allow a detailed analysis of the distribution error.

Regional wind monitoring system based on multiple sensor networks: A crowdsourcing preliminary test

The availability of updated information of regional wind fields is fundamental for an efficient management and prediction of wind power production. The present paper proposes a regional wind monitoring system based on the integration of different meteorological nets as a direct way to obtain this information. Concretely, we describe a monitoring system for the region of Andalusia supported by 198 stations, able to provide near surface wind field estimations with updating period of 10min. Each of these stations has different characteristics since they are focused on measuring different environmental parameters, and most of them are below the quality level required by the World Meteorological Organization for wind measurement. Despite this drawback, the proposed monitoring system takes advantage of the high density of measurement points to produce valuable descriptions of the regional wind field. A basic geostatistical model coupled to the system achieves better results than Numerical Weather Prediction models, obtaining RMSE values of 1.52m/s and 64.9° for speed and directional series respectively. The test confirms the tolerance of this massive monitoring system to data without quality assurance, being an interesting platform to implement crowdsourcing methodologies.

Performance of Mismatched PV Systems With Submodule Integrated Converters

Mismatch power losses in photovoltaic (PV) systems can be reduced by the use of distributed power electronics at the module or submodule level. This paper presents an experimentally validated numerical model that can be used to predict power production with distributed maximum power point tracking (DMPPT) down to the cell level. The model allows the investigations of different DMPPT architectures, as well as the impact of conversion efficiencies and power constraints. Results are presented for annual simulations of three representative partial shading scenarios and two scenarios where mismatches are due to aging over a period of 25 years. It is shown that DMPPT solutions that are based on submodule integrated converters offer 6.9-11.1% improvements in annual energy yield relative to a baseline centralized MPPT scenario.

Virtual Energy Information Network: a resilience perspective

Increasing demand in energy consumption, missed modernisations, and the increasing difficulties in predicting power production due to volatile renewable energy sources (e.g., based on wind or sun) impose major challenges to the power grid. Power supply and power demand are closely interconnected with the need to maintain the power grid in a stable state with a sufficient quality of power. This requires energy-relevant information to be exchanged through the so called Energy Information Network. Communication, however, is challenging within the Energy Information Network due to privacy, security, resiliency, and quality-of-service requirements. Particularly, the resilience of communication within the Energy Information Network needs to be considered to maintain the power grid in a stable and controlled state. This paper suggests a Virtualised Energy Information Network (VEIN), where the Energy Information Network is divided into multiple virtual networks that run over a common substrate network. Furthermore, this paper discusses benefits of this approach in terms of privacy, security, and resilience and points out open research questions.

Model of Photovoltaic Power Plants for Performance Analysis and Production Forecast

A photovoltaic (PV) plant model is presented. It is based on a detailed electrothermal description of the panels forming strings that, in turn, form the power plant. It accounts for environmental working conditions, such as temperature and wind speed, and specific plant configuration, such as plant topology and power losses due to interconnections. The input variables of the model are the ambient temperature, irradiance, and wind speed. The model derives the working temperature of the panel taking into account also the power conversion performed by the panel; the electrical operating point is determined by simulating the actions done by the maximum power point tracker that operates at plant level. This model has been tested using a large database of experimental data from industrial PV plants characterized by power levels ranging from 250 kW to 1 MW. As shown, the model is capable to predict power production when “fed” by forecast irradiance, ambient temperature, and wind speed data.

The impact of model physics on numerical wind forecasts

Fine scale numerical weather prediction (NWP) models are now widely applied to predict power production at wind farms. Given the fact that demand for specialized forecasts for wind farms is growing, it is important to understand the strengths and limitations of NWP models for producing wind forecasts. This paper seeks to partially fulfill this goal by exploring the sensitivity of NWP-based wind forecasts to the choice of model physics schemes. The authors used two distinct case studies to explore these sensitivities with a NWP model used in realtime wind power forecast, where the underlying meteorology in both cases had a profound impact on the wind ramp-up of a wind farm in Northern Colorado. The first case was a strong cold frontal system moving through the wind farm during winter, and the second case was for a line of strong thunderstorms passing through the wind farm during summer. The model results were compared with observed hub-height wind.In each case, sensitivity studies were conducted to explore the impact of the choice of physics schemes on the wind forecasts. For the winter case, the sensitivity to the representation of land surface and planetary boundary layer processes was quantified. For the summer case, the sensitivity to the representation of clouds and precipitation physics was explored. In the winter case, the wind forecast was less sensitive to the choice of physics schemes due to the hub-height winds being generated by strongly forced large-scale weather systems. In contrast, the wind forecast for the summer case (driven by weaker meteorological forcing) was strongly affected by the choice of cloud and precipitation physics schemes.

Calibration and validation of a model for simulating thermal and electric performance of an internal combustion engine-based micro-cogeneration device

The growing worldwide demand for more efficient and less polluting forms of energy production has led to a renewed interest in the use of micro-cogeneration technologies in the residential. Among the others technologies, internal combustion engine-based micro-cogeneration devices are a market-ready technology gaining an increasing appeal thanks to their high efficiency, fuel flexibility, low emissions, low noise and vibration.In order to explore and assess the feasibility of using internal combustion engine-based cogeneration systems in the residential sector, an accurate and practical simulation model that can be used to conduct sensitivity and what-if analyses is needed. A residential cogeneration device model has been developed within IEA/ECBCS Annex 42 and implemented into a number of building simulation programs. This model is potentially able to accurately predict the thermal and electrical outputs of the residential cogeneration devices, but it relies almost entirely on empirical data because the model specification uses experimental measurements contained within a performance map to represent the device specific performance characteristics coupled with thermally massive elements to characterize the device's dynamic thermal performance.At the Built Environment Control Laboratory of Seconda Università degli studi di Napoli, an AISIN SEIKI micro-cogeneration device based on natural gas fuelled reciprocating internal combustion engine is available. This unit has been intensively tested in order to calibrate and validate the Annex 42 model. This paper shows in detail the series of experiments conducted for the calibration activity and examines the validity of this model by contrasting simulation predictions to measurements derived by operating the system in electric load following control strategy. The statistical comparison was made both for the whole database and the segregated data by system mode operation.The good agreement found in the predictions of net electric power production, useful thermal output and primary power consumption allowed to conclude that the Annex 42 model can be used to carry out a detailed performance assessment in order to examine the applicability of the AISIN SEIKI unit for supplying building electrical and thermal energy requirements according to different load profiles during annual or multi-year operation.

Assessment of the benefits of numerical weather predictions in wind power forecasting based on statistical methods

Several forecast systems based on Artificial Neural Networks have been developed to predict power production of a wind farm located in a complex terrain, where geographical effects make wind speed predictions difficult) in different time horizons: 1,3,6,12 and 24 h. In the first system, the neural network has been used only as a statistic model based on time series of wind power; later it has been integrated with numerical weather predictions, by which an interesting improvement of the performance has been reached, especially with the longer time horizons. In particular, a sensitivity analysis has been carried out in order to find those numerical weather parameters with the best impact on the forecast. Then, after the implementation of forecast systems based on a single ANN, the two best prediction systems individuated through the sensitivity analysis, have been employed in a hybrid approach, made up of three different ANNs. Besides, a prediction system based on the wavelet decomposition technique has been also carried out in order to evaluate its contribute on the forecast performance in two time horizons (1 and 24 h). The error of the different forecast systems is investigated and the statistical distributions of the error are calculated and presented.

Adaptive Control of a Wind Turbine With Data Mining and Swarm Intelligence

The framework of adaptive control applied to a wind turbine is presented. The wind turbine is adaptively controlled to achieve a balance between two objectives, power maximization and minimization of the generator torque ramp rate. An optimization model is developed and solved with a linear weighted objective. The objective weights are autonomously adjusted based on the demand data and the predicted power production. Two simulation models are established to generate demand information. The wind power is predicted by a data-driven time-series model utilizing historical wind speed and generated power data. The power generated from the wind turbine is estimated by another model. Due to the intrinsic properties of the data-driven model and changing weights of the objective function, a particle swarm fuzzy algorithm is used to solve it.

The empirical validation of a model for simulating the thermal and electrical performance of fuel cell micro-cogeneration devices

Fuel cell micro-cogeneration is a nascent technology that can potentially reduce the energy consumption and environmental impact associated with serving building electrical and thermal demands. Accurately assessing these potential benefits and optimizing the integration of micro-cogeneration within buildings requires simulation methods that enable the integrated modelling of fuel cell micro-cogeneration devices with the thermal and electrical performance of the host building and other plant components. Such a model has recently been developed and implemented into a number of building simulation programs as part of an International Energy Agency research project. To date, the model has been calibrated (tuned) for one particular prototype 2.8 kW AC solid-oxide fuel cell (SOFC) micro-cogeneration device. The current paper examines the validity of this model by contrasting simulation predictions to measurements from the 2.8 kW AC prototype device. Good agreement was found in the predictions of DC power production, the rate of fuel consumption, and energy conversion efficiencies. Although there was greater deviation between simulation predictions and measurements in the predictions of useful thermal output, acceptable agreement was found within the uncertainty of the model and the measurements. It is concluded that the form of the mathematical model can accurately represents the performance of SOFC micro-cogeneration devices and that detailed performance assessments can now be performed with the calibrated model to examine the applicability of the 2.8 kW AC prototype device for supplying building electrical and thermal energy requirements.