Encouraging households with PV panels to consume more of their self-generated electricity

How to cite

Akesson, Jesper; Metcalfe, Robert; Kacha, Ondrej. 2025. "Encouraging households with PV panels to consume more of their self-generated electricity." AEA RCT Registry. AEARCTR-0015605. https://doi.org/10.1257/rct.15605-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Jesper Akesson* , Robert Metcalfe , Ondrej Kacha

* Primary investigator

Registry ID
AEARCTR-0015605
DOI
https://doi.org/10.1257/rct.15605-1.0
First registered
Last updated
Status
In development
Trial dates
2025-01-01 – 2025-10-30
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/15605

Abstract

Households with installed solar panels have significant potential to increase residential demand flexibility. By generating their own electricity, solar-powered households can reduce their dependence on the grid. However, they also pose a challenge - that is they sell excess solar power back to the grid during off-peak hours (typically between 10am - 4pm on week days) when electricity supply already exceeds demand. This surplus can negatively impact grid stability. A key way to mitigate this risk is to increase the capability and motivation of households to increase the self-consumption of their solar power during off-peak periods.

The goal of this study is to test whether providing households with a "solar efficiency score" (i.e., a comparison of self-consumed solar energy vs. solar energy fed to the grid) can increase their self-consumption of solar energy..

Experimental design

We will randomise approximately 25,000 customers into two trial groups (5,000 customers in control and 20,000 customers in treatment group). Customers in the treatment group receive a behavioural email which encourages them to view their solar efficiency score, and to sign up for a newsletter with tips to improve their solar efficiency. The aim is to reach at least 5,000 newsletter sign ups. If the sign-up target is not reached, the energy supplier will seek to increase the sample size so that the target is reached.

The behavioural email will be sent to customers in April 2025. The email will contain a link to a website that calculates a solar efficiency score based on the number of installed solar panels (inputted manually by a consumer) and last month’s energy feed-in data (loaded automatically for each user). Finally, upon presentation of the solar efficiency score on the website, customers will be given an option to sign up for three emails sent over the next three months that contain tips to improve their solar efficiency score.

Using customers' daily solar feed in and hourly energy usage data for three months before and six months after the intervention, we will evaluate the impact of viewing solar efficiency score and receiving emails with tips on solar energy feed-in and energy consumption (peak, offpeak, and overall). The effect of this campaign will be tested using an instrumental variable regression. We use the difference in enrollment caused by our treatment allocation as an instrument to understand the causal impact of the campaign on solar feed-in and energy consumption.

Design details

Randomization method
Randomization done in office by a computer
Randomization unit
Household
Clusters
25,000 households
Observations
25,000 households
Arms
5,000 households control, 20,000 households treatment
Primary outcomes
energy consumption, energy feed-in

Cited by

No citation data available for this registration yet.

Keywords