Mapping coal vs. solar and wind electricity costs in the U.S.
Working with two friends at MIT, I developed an online tool, CoalMap, to help activists, regulators, and the general public explore the economic costs of existing coal-fired power plants in the United States. Drawing from publicly available datasets, we identified U.S. coal plants that are particularly vulnerable to shutdown efforts by comparing each plant’s average operating cost to the levelized cost of electricity (LCOE) for new utility-scale solar photovoltaic (PV) and wind generation in the same location. Users can apply different carbon prices and rates of cost decline for solar and wind, and observe the effects on the cost-competitiveness of renewable generation in future years.
We describe the findings of our analysis in a MIT Energy Initiative working paper released in early 2016. Our findings highlight the importance of sustained technology improvement and appropriate public policy in shifting the U.S. electricity generation mix toward low-carbon sources, and ultimately in achieving national climate-change mitigation targets. With 5% annual renewable cost declines and a carbon price equal to the U.S. government’s social cost of carbon, new unsubsidized wind and solar PV generation at existing coal plant locations will be cost-competitive with fully amortized U.S. coal plants by 2019 and 2031, respectively.
- Winner of 2015 MIT Clean Earth Hackathon
- J. Jean, D.C. Borrelli, and T. Wu, "Mapping the economics of U.S. coal power and the rise of renewables," MITEI Working Paper WP-2016-01 (2016).
- Press highlights: MIT News (1, 2)
- CoalMap has been used by the Rocky Mountain Institute, Sierra Club, and the Environmental Defense Fund in their advocacy efforts to move the U.S. power system beyond coal.
Solar USB Phone Charger
Powered by sunlight. Works for tablets too.
With Fossil Free MIT, I helped develop a simple and low-cost solar-powered USB charger as a hands-on engineering activity for kids and adults of all ages. Anyone can make their own at home in about 2 hours with just a soldering iron and a few basic tools. The charger has a regulated 5V USB output and produces a photocurrent of ~1.5A under peak sunlight, comparable to the nominal 5V/1A output of a typical iPhone 5s charger. Our procedure is posted below. Feel free to contact me with questions or suggested improvements!
Procedure for building a solar-powered USB charger
Han Solar: A Short Film on Thin-Film Solar Cells
An educational video produced with the Museum of Science
In January 2014, I participated in a weeklong internship at the Museum of Science (Boston) via the MIT Center for Excitonics Research Communication Laboratory. As an exercise in science communication and educational media, I worked with Sarah Luppino from the Swager Lab (MIT Chemistry) to produce a stop-action video explaining how thin-film solar cells are made in our own labs at MIT.
Many thanks to Carol Lynn Alpert, Karine Thate, and Jeanne Antill of the Museum of Science for their invaluable guidance in the production and editing of this video.
Rooftop Solar Tracking
A low-cost solar tracker for residential solar PV systems
A solar harvester captures the most energy when pointed directly at the sun. Solar panels and sunflowers alike can track the sun from east to west to make the most of the limited solar flux at the earth's surface. But solar tracking only makes sense if it generates more energy than it consumes. Practically speaking, that extra energy has to be worth more than the tracker itself costs. Making a cheap solar tracker that tracks accurately, looks good, and survives wind, snow, and bird poop can be an engineering nightmare. That's why only ~30% of utility-scale PV systems—and nearly no household installations—employ tracking today.
In fall 2013, I worked with 3 undergrad mechanical engineers (Claire Kearns-McCoy, Jordan Mizerak, and Manuel Romero) to design a solar tracking system for PV arrays on inclined residential rooftops, as part of a MIT class on medical device design* taught by Alex Slocum and Charlie Sodini. The National Renewable Energy Lab (NREL) offers free software tools (see PVWatts and SAM) for assessing grid-connected renewable energy projects. To do a more detailed analysis, we developed our own computational model. The model predicts the annual PV energy output gain and cost constraints for different solar tracking strategies. Since it draws on publicly available insolation and meteorological data from NREL, this analysis can be generalized to any location in the U.S.
Our analysis suggests that tracking on inclined rooftops can be economically viable, even in Boston. With computational results in hand, we designed and built a prototype one-axis rooftop solar tracker based on low-cost components (Arduino microcontroller, accelerometer, DC windshield wiper motor, and bits of 80/20). The result? I wouldn't invest yet, but the concept is promising and the problem important enough to be worth a second prototype. For me and my teammates, it was a learning experience like no other.
*Are solar panels medical devices? Maybe not in the conventional sense. But by replacing highly polluting carbon-based fuels like coal and natural gas, solar and other renewable energy technologies could help eliminate a major threat to public health.
Grid-Scale Energy Storage
An analysis of grid energy storage requirements for high penetration of intermittent renewables
Achieving climate mitigation targets will require a global shift from existing fossil-driven electricity generation to low-carbon renewable generation. But not all electricity is created equal: As demand (or load) varies throughout the day and throughout the year, supply must follow. Load-following capability, also known as dispatchability, is a natural advantage of conventional thermal power plants. Gas turbines can be turned on or off within minutes, and coal-fired units can vary their output throughout the day by adjusting fuel consumption. Renewable sources, however, have no such levers. Wind turbines and solar cells generate electricity only when the wind blows and when the sun shines. While their total resource is sufficient to meet world electricity demand many times over, the availability of wind and solar energy varies on seasonal and daily time scales, as well as from hour to hour and minute to minute based on local meteorological conditions. This variability is referred to as intermittency.
Intermittency and the resulting temporal mismatch between resource availability and electricity demand limit the maximum penetration of grid-connected renewable generation. Energy storage can increase renewable penetration by mitigating intermittency and providing ancillary services to the grid, including voltage regulation, frequency regulation, and power conditioning. The cost and performance of storage technologies are improving quickly, but deployment of grid-scale storage remains low. Key obstacles to deployment include high capital costs, limited charge and discharge rates (power capacity), and limited total energy capacity. Existing grid-scale installations are mostly pumped hydroelectric storage (PHS, or pumped hydro) systems, which account for ~99% of global installed storage capacity.
In Prof. Jessika Trancik's course on energy systems and climate change mitigation, I worked with two students (Nora Xu and Joshua Mueller) to develop a simulation-based model of an electric grid in which solar PV is the only available generation source, with varying levels of energy storage. This model shows how storage system performance parameters affect the amount of electricity demand that cannot be satisfied by PV. The remaining demand must then be met by peaker plants and other dispatchable generation.
We estimated electricity supply and demand from insolation and regional load data, assuming different levels of installed PV capacity, represented as area covered by PV farms. Taking North Central Texas (ERCOT) as a case study, we determined optimal threshold energy and power capacities for minimizing the severity and frequency of energy deficits. Our findings provide insight into the relative importance of storage performance metrics for mitigating the intermittency of renewable energy generation. The conclusion? All the energy storage in the world today is hardly enough to meet the power capacity needs of a 100% solar-powered Texas. We need to get to work.