Public policies that aim to reduce carbon emissions act as a constraint on production decisions. Much of the empirical literature has shown that environmental policies reduce carbon emissions, but the source of that reduction is not explained. This paper investigates the hypothesis that environmental policy directly affects production decisions. Hence, we look at the energy mix of firms, emphasizing the choice firms face when deciding which energy input to use. We build a difference-indifferences model to show how the implementation of the first three phases of the European carbon market (that is, the EU Emissions Trading System) influenced the composition of the energy mix in favor of carbon-neutral energy inputs (ie, firms substitute between energy inputs in favor of less carbon-intensive inputs). To do this, we use French microdata on manufacturing firms, specifically information on energy usage, carbon emissions and assignment to treatment. We show that when considering the energy mix, firms that were subject to the European carbon market tended to use more carbon-neutral energy inputs during the three phases of the policy, with the magnitude varying according to the specification.

In this paper we extend an existing Markov regime-switching model approach for power spot prices by combining it with a seasonal autoregressive moving average model, which allows us to incorporate intraday and weekly seasonalities. In addition, we also put forward top-down models for wind power, solar power and power load. Moreover, we present a bottom-up model for the hourly power yield of a solar panel. Finally, we exploit the aforementioned models to simulate the hourly price and solar panel power yield paths in order to estimate the spot market risk of a small hypothetical solar array in the context of a power purchase agreement (PPA). PPAs allow the risk transfer from renewable plant operators to more trading and risk management skilled market participants. Spot market risk arises because neither the volume nor the exact power spot prices are known when PPAs are concluded.

The efficient market hypothesis (EMH) is a fundamental tenet of active and unfettered market trading. The EMH for intertemporal trading of a commodity such as natural gas implies that today’s futures price is an unbiased predictor of a future delivery period’s spot price and that the difference between today’s futures and spot prices reflects the commodity’s cost of carrying. The EMH for inter-regional trading requires the commodity’s regional price difference to equal the inter-regional transportation cost. Using regional hourly data for January 1, 2011 to December 31, 2020 from the Electric Reliability Council of Texas (ERCOT), we investigate whether the EMH is empirically valid in all ERCOT’s wholesale electricity markets and, if not, the extent of ERCOT’s energy trading inefficiency and what can be done to reduce it. By estimating a parsimonious system of eight price-level regressions and four price-difference regressions, we reject the interday EMH for all regions and the inter-regional EMH for all regional market pairs. However, ERCOT’s extent of trading inefficiency is mild when compared to the wholesale energy prices. Our empirical examples have two policy implications: enhancing ERCOT’s interday trading efficiency entails improving the day-ahead forecasts for solar and wind generation and refining ERCOT’s indicative real-time market prices, while enhancing ERCOT’s inter-regional trading efficiency requires transmission capacity expansion to reduce transmission congestion and line losses.

This paper offers an insight into the infrastructure of hydrogen production in order to achieve a clear vision for electrolyzer planning and to avoid the possibility of huge sunk costs at an early stage. By means of a power plant dispatching model, we evaluate the potential for hydrogen production in France using the residual load of renewables and nuclear power after generating sufficient power to meet the electricity demand. Our results show that intermittency in generation is passed on through electrolyzers, leading to low capacity factors and unclear business models. Installing decentralized multiple electrolyzers with learning rates seems a better strategy than installing a few large electrolyzers with economies of scale; the aim is to achieve this strategy by 2035, with the trend reversed thereafter. However, the cost of hydrogen is set to increase by 2050 due to low capacity and low residual load. Thus, the massive deployment of clean hydrogen production requires ambitious power generation capacities, implying that connecting hydrogen only to surplus power-in might not be enough to decarbonize economies. Dedicated wind and solar power farms or imports of low-carbon hydrogen will most probably be necessary.