• Energy Research
• 2021-2025close
• Closed Access close
• Restricted close
• Embargo close
• Energy close

Abstract Accurate power forecasting is of great importance to the turbine control and predictive maintenance. However, traditional physics models and statistical models can no longer meet the needs of precision and flexibility when thermal power plants frequently undertake more and more peak and frequency modulation tasks. In this study, the recurrent neural network (RNN) and convolutional neural network (CNN) for power prediction are proposed, and are applied to predict real-time power of turbine based on DCS data (recorded for 719 days) from a power plant. In addition, the performances of two deep learning models and five typical machine learning models are compared, including prediction deviation, variance and time cost. It is found that deep learning models outperform other shallow models and RNN model performs best in balancing the accuracy-efficient trade-off for power prediction (the relative prediction error of 99.76% samples is less than 1% in all load range for test 216 days). Moreover, the influence of training size and input time-steps on the performance of RNN model is also explored. The model can achieve remarkable performance by learning only 30% samples (about 216 days) with 3 input time-steps (about 60 s). Those results of the proposed models based on deep-learning methods indicated that deep learning is of great help to improve the accuracy of turbine power prediction. It is therefore convinced that those models have a high potential for turbine control and predictable maintenance in actual industrial scenarios.

A hybrid-twin for gas-bearing supported, high-speed turbocompressor is suggested to be an aide to increase the operational reliability and to provide valuable insights to improve its design. The bearing clearances of a few micrometers imply that excessive thermal or mechanical deformations can result in machine failure due to mechanical seizure or rotor-dynamic instabilities. The high centrifugal forces and thermal gradients may induce material fracture or the lift-off of press-fitted assemblies. The mentioned phenomena and interactions depend strongly on the operating conditions and imply a risk of hitting critical operating zones potentially unscreened during the design phase. The developed twin asset involves 1D multi-domain models that are accurate enough to provide useful information regarding the most critical interactions and are computationally viable for real-time applications suggesting a run-to-real time ratio of 2 per cent. A case study in which a turbocompressor touchdown is analyzed a posteriori using the twin asset to highlight the possible insights to be gained by using the virtual twin to predict data that cannot be readily measured such as the thrust bearing axial clearance under the impeller thrust force. Finally, the generation of signals by the twin asset providing a degree of redundancy with sensor values is highlighted as an opportunity to increase confidence in monitoring tasks, paving the way to extension into control strategies.

Abstract China's carbon emissions have been ranking first in the world. This study filled in the gaps in research, decomposed carbon intensity from the perspective of time, space and industry. A decoupling effort model based on factor decomposition models was constructed to analyze the driving factors of carbon emissions and economic decoupling, which builded a foundation for achieving sustainable economic development. Using the Logarithmic Mean Divisia Index method (LMDI), the paper measured the carbon emission intensity of 29 provinces and cities in China from 1998 to 2019, and decomposed the decoupling effect between GDP and carbon emission on the basis of factor decomposition by tapio. The results showed that: (1) Carbon intensity declined first, then rise lightly, and finally declined steadily. For the primary industry and the tertiary industry, the carbon intensity declined steadily, while the carbon intensity increased accordingly to the overall carbon intensity. In terms of spatial evolution, the regional differences between different provinces decreased correspondingly. (2) The cumulative contribution rates of these three effects, i.e., technological progress, industrial structure and regional scale were 106.3299%, −15.1486% and 8.8188%, respectively. There were obvious differences of these cumulative contribution rates of carbon intensity among different provinces. (3) From the perspective of industrial, technological progress effect is the largest contribution for carbon intensity in the secondary industry. The Industrial structure effect mainly affects the primary and tertiary industries; and no significant difference in regional scale effect. (4) The decoupling effect gradually improved, and technological progress has played an absolute leading role in promoting the decoupling effect. Based on the research results, the key policy recommendation are put forward as follows: (1) Further improve the technological level and support clean technology enterprises. (2) Promote industrial upgrading in backward industrial provinces (3) Promote regional assistance and the introduction of high-quality foreign investment.

Abstract Due to gradually diminution of fossil fuel, the cost-effective utilization of available fuel for power generation has turn out to be a vital concern of electric power utilities. Thermal power plants have to operate within their fuel confines and contractual constraints. This work suggests social group entropy optimization (SGEO) technique to solve short-term generation scheduling of a power system consisting of fuel constrained thermal generating units, cascaded hydro power plants, solar PV plants, wind turbine generators and pumped storage hydro (PSH) plants with demand side management (DSM). Simulation results of the test system have been compared with those acquired by self-organizing hierarchical particle swarm optimizer with time-varying acceleration coefficients (HPSO-TVAC), fast convergence evolutionary programming (FCEP) and differential evolution (DE). Numerical results show that fuel consumption can be adequately controlled for fulfilling constraints imposed by suppliers and total cost with fuel constraints is more than the cost without fuel constraints. It has been also observed from the comparison that the suggested SGEO has the ability to bestow with superior-quality solution.

Abstract Propane (C3H8) is being considered as an alternative refrigerant, besides being used as an alternative fuel, because of its low Global Warming Potential and zero Ozone Depletion Potential. Using blends of C3H8 with CO2 as refrigerants, diminishes the fire safety concerns in case of accidental leak of this flammable substance in refrigeration applications. This paper reports on the effects of CO2 on laminar burning speed and flame instability of C3H8/air blends at elevated temperatures and pressures. The flame structures were investigated in a Schlieren system. The laminar burning speeds of C3H8/CO2/air mixtures were measured in a spherical chamber and were fitted by a power-law mathematical correlation. The one-dimensional flame code from Cantera with kinetic model was also used to predict laminar burning speed. The high-speed photography showed that CO2 inhibits the flame instability because of its hydrodynamic and diffusional-thermal effects. Results showed that the laminar burning speed decreased with increasing CO2 mole fraction in the mixtures and that CO2 promotes the flame stability. The high temperature C3H8 oxidation was governed by the reaction of H + O 2 = O + OH . The effects of CO2 on laminar burning speed were mainly determined by the reaction of CO 2 + H = CO + OH and the high energy capacity (specific heat) of CO2.

Abstract Owing to the high penetration of renewable energy sources, power systems require more flexibility to respond to the fluctuation of variable renewable energy sources. However, flexibility provision by distributed energy resources (DERs) installed on the demand side is a cost-effective measure. In this study, we proposed a scheme that utilizes time-of-use (TOU) electricity rates as an incentive to encourage residential prosumers to provide more flexibility by controlling the DERs. In particular, we developed an energy management system model that applies model predictive control. The TOU rates were designed by an aggregator who traded with markets and prosumers. The DERs of residential prosumers were fuel cells with combined heat and power, CO2 heat pump water heaters, or batteries. Numerical simulations using measured energy demand and insolation data were performed to evaluate the performance of the developed model. According to the results, the developed model could integrate prediction errors and provide flexibility to the grid. Finally, the economies of the aggregator and prosumers were also evaluated.