Unprecedented high volumes of data are becoming available with the growth of the advanced metering infrastructure. These are expected to benefit planning and operation of the future power system, and to help the customers transition from a passive to an active role. In this paper, we explore for the first time in the smart grid context the benefits of using Deep Reinforcement Learning, a hybrid type of methods that combines Reinforcement Learning with Deep Learning, to perform on-line optimization of schedules for building energy management systems. The learning procedure was explored using two methods, Deep Q-learning and Deep Policy Gradient, both of them being extended to perform multiple actions simultaneously. The proposed approach was validated on the large-scale Pecan Street Inc. database. This highly-dimensional database includes information about photovoltaic power generation, electric vehicles as well as buildings appliances. Moreover, these on-line energy scheduling strategies could be used to provide realtime feedback to consumers to encourage more efficient use of electricity.

As the need for alternative transportation fuels increases, it is important to understand the many effects of introducing fuels based upon feedstocks other than petroleum. Water intensity in "gallons of water per mile traveled" is one method to measure these effects on the consumer level. In this paper we investigate the water intensity for light duty vehicle (LDV) travel using selected fuels based upon petroleum, natural gas, unconventional fossil fuels, hydrogen, electricity, and two biofuels (ethanol from corn and biodiesel from soy). Fuels more directly derived from fossil fuels are less water intensive than those derived either indirectly from fossil fuels (e.g., through electricity generation) or directly from biomass. The lowest water consumptive (<0.15 gal H 2 O/mile) and withdrawal (<1 gal H 2 O/mile) rates are for LDVs using conventional petroleumbased gasoline and diesel, nonirrigated biofuels, hydrogen derived from methane or electrolysis via nonthermal renewable electricity, and electricity derived from nonthermal renewable sources. LDVs running on electricity and hydrogen derived from the aggregate U.S. grid (heavily based upon fossil fuel and nuclear steam-electric power generation) withdraw 5-20 times and consume nearly 2-5 times more water than by using petroleum gasoline. The water intensities (gal H 2 O/mile) of LDVs operating on biofuels derived from crops irrigated in the United States at average rates is 28 and 36 for corn ethanol (E85) for consumption and withdrawal, respectively. For soyderived biodiesel the average consumption and withdrawal rates are 8 and 10 gal H 2 O/mile.

This work estimates the energy embedded in wasted food annually in the United States. We calculated the energy intensity of food production from agriculture, transportation, processing, food sales, storage, and preparation for 2007 as 8080 ± 760 trillion BTU. In 1995 approximately 27% of edible food was wasted. Synthesizing these food loss figures with our estimate of energy consumption for different food categories and food production steps, while normalizing for different production volumes, shows that 2030 ± 160 trillion BTU of energy were embedded in wasted food in 2007. The energy embedded in wasted food represents approximately 2% of annual energy consumption in the United States, which is substantial when compared to other energy conservation and production proposals. To improve this analysis, nationwide estimates of food waste and an updated estimate for the energy required to produce food for U.S. consumption would be valuable.

This letter consists of a first-order analysis of the primary energy embedded in water in the United States. Using a combination of top-down sectoral assessments of energy use together with a bottom-up allocation of energy-for-water on a component-wise and service-specific level, our analysis concludes that energy use in the residential, commercial, industrial and power sectors for direct water and steam services was approximately 12.3 ± 0.3 quadrillion BTUs or 12.6% of the 2010 annual primary energy consumption in the United States. Additional energy was used to generate steam for indirect process heating, space heating and electricity generation.