The solar rebound effect in residential PV

Researchers from Australia and Vietnam have analyzed the short and long-term dynamics of the so-called solar energy rebound effect.

The rebound effect consists of a reduction in expected gains from a more resource-efficient technology as a result of behavioral or systemic change. As for residential solar, this effect equals the change in a household's electricity consumption resulting from PV power generation.

“The issue occurs in almost every country, no matter whether it is developed or developing. The issue we found in Vietnam, of which customers abandoned solar systems, is a new finding. However, we believe this could also happen in any country where the policy supporting solar power is not fully and scientifically explained, and where consumers are not comprehensively consulted,” the research's corresponding author, Luan Nguyen, told pv magazine. 

The research group used a panel dataset encompassing approximately 3500 households in Hanoi, Vietnam, and obtained two datasets on the household and district level, respectively, covering the period 2015-2021. The panel dataset contains detailed monthly information on electricity demand and bills, solar system capacity, production, and installation time. About 48% of the households had PV, and the rest used electricity solely from the grid.

“If households that adopt solar panels also substantially change their energy consumption patterns, it is possible that the overall carbon emissions associated with energy use may be relatively unchanged. This phenomenon is known as a ‘rebound effect,’ where increased supply fails to induce substitution between solar and grid energy,” the scientists explained. “Notably, all homes in the dataset did not possess solar batteries, but none fed solar energy into the national grid. It means all solar owners either consumed all solar production or wasted it.”

To estimate the effects of PV installations on consumption, the researchers used the difference-in-differences identification strategy, which is a statistical method used to calculate the causal impact of an intervention by comparing it to a group that did not receive the same treatment.

“Leveraging a substantial, distinctive, and up-to-date panel dataset, we unveil a substantial rebound effect that challenges the presumed environmental advantages of residential solar energy,” the researchers highlighted. “We are also (to our knowledge), among the first authors to study the dynamics of this process and show that both the primary effect and the rebound effect tend to diminish over time.”

Based on their data, the academics extracted four estimation outputs: two for households’ total electricity consumption and two for demand from the grid. “The only differences between models in each pair are the inclusion or exclusion of exogenous variables, and those differences are to test the consistency of the estimations,” they explained. Exogenous variables included electricity demand, price, income, and other social parameters.

The academics found that, when including exogenous variables, solar installations immediately increased the total electricity demand by approximately 16.3% compared to non-solar households. However, it diminished to about 3.5% after 13 months. Also, the consumption from the grid fell by 3.6% after installation, fading out to about 1.5% at the end of the period.

“Since households that install solar panels may possess unobserved characteristics that distinguish them from households that did not, there is the potential for our results to be affected by unobserved confounding,” they stressed. “However, across a range of diagnostic methods, we find little evidence that endogeneity or misspecification issues may be biasing our results.”

The research group comprised scientists from Australia’s Griffith University, the University of Adelaide, and Vietnam’s Hanoi Power Corporation. Their findings were introduced in the paper “Solar rebound effects: Short and long term dynamics,” published on Renewable Energy.