In a field experiment, we assess the effectiveness of time-of-use tariffs in shifting intraday electricity demand toward the period when most production of electricity from solar sources takes place. Two treatment groups are created. The first treatment group faces a two-part tariff, with a low tariff between 11am and 3pm (the period with most solar radiation) and a high tariff outside these hours. The second treatment group faces a dynamic tariff, where periods of low and high tariffs change from day to day, according to day-ahead weather forecasts. The low tariff hours are transmitted to households one day in advance via text message. The control group faces a flat tariff. All households are equipped with smart meters that transmit electricity usage every hour (or even every 15 minutes). The experiment will last for at least one year.