Long-Term Projection of Renewable Energy Technology Diffusion

Abstract

The EU aims at increasing the use of renewable energy sources (RES), mainly solar-photovoltaic (PV) and wind technologies. Projecting the future, in this respect, requires a long-term energy modeling which includes a rate of diffusion of novel technologies into the market and the prediction of their costs. The aim of this article has been to project the pace at which RES technologies diffused in the past or may diffuse in the future across the power sector. This analysis of the dynamics of technologies historically as well as in modeling, roadmaps and scenarios consists in a consistent analysis of the main parameters of the dynamics (pace of diffusion and extent of diffusion in particular markets). Some scenarios (REMIND, WITCH, WEO, PRIMES) of the development of the selected power generation technologies in the EU till 2050 are compared. Depending on the data available, the learning curves describing the expected development of PV and wind technologies till 2100 have been modeled. The learning curves have been presented as a unit cost of the power versus cumulative installed capacity (market size). As the production capacity increases, the cost per unit is reduced thanks to learning how to streamline the manufacturing process. Complimentary to these learning curves, logistic S-shape functions have been used to describe technology diffusion. PV and wind generation technologies for the EU have been estimated in time domain till 2100. The doubts whether learning curves are a proper method of representing technological change due to various uncertainties have been discussed. A critical analysis of effects of the commonly applied models for a long-term energy projection (REMIND, WITCH) use has been conducted. It has been observed that for the EU the analyzed models, despite differences in the target saturation levels, predict stagnation in the development of PV and wind technologies from around 2040. Key results of the analysis are new insights into the plausibility of future deployment scenarios in different sectors, informed by the analysis of historical dynamics of technology diffusion, using to the extent possible consistent metrics.

Impact of the Lockdown during the COVID-19 Pandemic on Electricity Use by Residential Users

“…Distributed resources, located close to energy users, can play an important role here. In connection with the development and diffusion of RES technologies [63], photovoltaic (PV) sources seem to be particularly predestined to meet the increased demand for energy during the daytime hours by residential consumers. It is possible to develop prosumer PV installations on the roofs and facades of multi-family buildings (e.g., [64][65][66][67], including green roofs [68]).…”

Digitalization in the Renewable Energy Sector

“…The largest share of this RES contribution (25%) will be represented by wind and solar PV sources combined [34]. From an environmental perspective, by 2050 Europe alone is projected to have over 450 GW capacity in wind power, with avoided emissions of 462 Mt CO 2 per year, and 4500 TWh capacity in solar photovoltaics (PV), with avoided emissions of 2.3 Gt of CO 2 per year [35].…”

The technology life cycle of Persian lime. A patent based analysis

“…Lastly, Skoczkowski et al. [ 66 ], did research on how technologies like solar-photovoltaic and wind power spread. .…”

“…It is common in the literature to find research with complementary methods to the technology life cycle that improves the prediction or analysis of the technology or sector. Thus, some research aims to link not only the innovation approach or the technology diffusion, but the market size [ 66 ] or stock market behaviour over time [ 15 ] using patents on one side and variables such as unit of cost per installed capacity or the NASDAQ composite stock index on the other. Also, research uses complementary management tools to the S-curve such as technology roadmaps and technology audits [ 80 ], and specific methods like machine learning [ 85 ], the use of entropy [ 77 ], the Dirichlet allocation topic model [ 81 ], the use of time series modelling using statistical quality control charts [ 79 ], or hierarchical S-curves to analyse main technologies (e.g.…”