Googling grad students

August 30, 2013

I googled “grad student” along with a number of different search terms. Here are the results, scientifically speaking:

Google grad student

The lesson here isn’t that you shouldn’t go to grad school… Just make sure you always use a log scale.

Edit (20130903): This plot was featured on PhD Comics! http://www.phdcomics.com/comics.php?f=1626.


The Things They Carried: All aboard the Carnival Paradise

August 22, 2013

Last weekend’s Caribbean cruise with my brother and friends left me with some lessons in packing, a boatload (ha ha) of memories, and a few supremely entertaining photos.

For Future Me and anyone else planning and packing for a cruise, here’s a list of items you won’t remember to bring but might turn out to be critical (by some hedonistic criteria):

  • Water and drinks – Only lemonade and iced tea are provided. If you want anything to drink besides ship tap water without paying exorbitant fees, bring it with you.
  • Alcohol – Ditto. Carnival allows each person to bring one bottle of wine onto the ship; everything else is confiscated UNLESS you’re sneaky. Contact Robin Tan for tips on how to be sneaky.
  • Cash – Always good to have. Especially if you like to gamble. (Please gamble responsibly.)
  • Point-and-shoot camera – Cell phones don’t work on a ship for ordinary people on ordinary budgets, so there’s no reason to carry one around. Bring a dedicated camera for higher-quality memories.
  • Walkie-talkies – How else are you going to coordinate that 19th group outing to the 24-hour pizza buffet on Lido deck?
  • Games – There’s always time for games. Too much time.
  • Speakers – Just in case you want to listen to something besides that old dude singing karaoke.
  • Gym clothes – You’ll need to work out hard to power through that fourth appetizer at dinner.
On to the photos: The best of the bunch are simply too awesome for public consumption and will be left to the imagination out of respect for the subjects’ careers and dignity, but here are a few solid runners-up.
This guy is cooler than you.

This guy is cooler than you.

This guy just thinks he is.

This guy just thinks he is.

Sometimes you lose money at the casino. Sometimes Carlina wins $1000 in cold, hard cash.

Sometimes you lose money at the casino. Sometimes Carlina wins $1000 in cold, hard cash.

Beka and Robin only won $200. Weak.

Beka and Robin only won $200. Weak.

Offloading ourselves in Cozumel.

Offloading ourselves in Cozumel.

Candid shot.

Candid shot.

Looking for a sunken ship during the Amazing Race in Cozumel.

Looking for a sunken ship during the Amazing Race in Cozumel.

We win!

We win!

Gold medals!

Gold medals!

At the beach.

At the beach.

The first in a series of pool escapades.

The first in a series of pool escapades.

Human totem pole, AKA Double Chicken.

Human totem pole, AKA Double Chicken.

Human pyramid. Why are the girls always on top?

Human pyramid. Why are the girls always on top?

Again.

Again.

Last hurrah in Cozumel.

Last hurrah in Cozumel.

Edible happiness at breakfast.

Edible happiness at breakfast.

Birthdays happen at sea too. Happy 24th and 25th, Carlina and Justin!

Birthdays happen at sea too. Happy 24th and 25th, Carlina and Justin!

The boys.

The boys.

The Duke girls.

The Duke girls.

On a post-cruise diet in Tampa.

On a post-cruise diet in Tampa. Thanks John and Robin!

Team pic: Carlina, me, Robin, Neal, Justin, Dave, Beka, Nancy. Thanks for a ridiculously awesome weekend!

Team pic: Carlina, me, Robin, Neal, Justin, Dave, Beka, and Nancy. Thanks for a ridiculously awesome weekend!


It’s just the future of the world. No big deal.

July 30, 2013

Never would I have thought that something as distant and mundane as the climate could upset me so.

But I’ve been pondering climate change more and more with every passing week, and more and more it’s becoming clear that we (humans) have two options if we want to avoid the icky consequences of climate change:

  1. Stop burning fossil fuels for energy.
  2. Find a new planet to live on.

Tongue-in-cheekiness aside, I truly do believe in Option 2—space exploration—as a vital long-term goal for humanity. Unfortunately, the time for climate action is limited, and new worlds aren’t built in a day. So today I thought it would be a good idea to link climate change with my own interests and background in energy, to summarize and share some of the unadulterated facts. I hope to inspire you to take a moment to think, to personalize and internalize the numbers, to dedicate your own prodigious intellectual resources to finding a solution to our shared little problem.

I draw numbers and inspiration (and a few direct quotes) from a 2009 review paper by Stanford’s Mark Jacobson, whose work I highly recommend to anyone interested in energy and the environment.


Climate Change and Air Pollution

Key point: Climate change and air pollution have a host of deleterious effects on human and environmental health.

Climate change (or global warming) is caused by an increase in the greenhouse effect induced by the following human inputs into the atmosphere (from most to least significant):
  • CO2
  • Fossil and biofuel soot
  • Methane
  • Halocarbons
  • Tropospheric ozone
  • Nitrous oxide

So what? Well, here’s my own little climate-change mnemonic:

The effects of climate change are FAST.

  • Food and Freshwater
    • Shifts viable locations for agriculture
    • Shifts the timing and magnitude of freshwater availability
  • Animals and plants, Acidity
    • Alters natural habitats substantially, causing extinction of plant and animal species
    • Increases the acidity of the ocean (dissolved CO2 forms carbonic acid, H2CO3) – The pH of the world’s oceans has dropped by ~0.1 since pre-industrial levels
  • Storms, Sea levels, and Sickness
    • Makes tropical storms more severe
    • Shrinks glaciers, snow pack, and ice levels worldwide, causing sea levels to rise
    • Increases the prevalence or geographic spread of various diseases (e.g., West Nile)
  • Temperature
    • Increases global temperatures and accompanying heat-related health risks

Air pollution is the 6th leading cause of death worldwide; it has been implicated by the WHO in 2.4 million deaths worldwide each year. Indoor and outdoor pollutants can increase ER visits, reduce worker productivity, and induce all of the following health conditions: asthma, respiratory illness, cancer, and cardiovascular disease.

Slightly troubling note: Human-created aerosol particles in the atmosphere—sulfates, nitrates, chlorides, ammonium, potassium, carbon, H2O—mask ~50% of the current global warming potential; if we reduce air pollution, much of that warming potential will be unveiled.


Energy Production and Use

Key point: Both climate change and air pollution are fundamentally linked to energy production and use; both arise from the burning of solid, liquid, and gaseous fuels. Solving these problems will thus require fundamental changes to the energy sector.

A few important numbers on energy and emissions:

  • Global energy production = 133 PWh/yr = 15.2 TW (2005)
  • Global electricity production = 18.2 PWh/yr = 2.1 TW (2005)
  • 25% of total US CO2 emissions are directly exhausted from vehicles on the road; another 8% are due to production and transport of fuel. Operation of fossil-fuel-burning vehicles thus accounts for over 33% of total US CO2 emissions, which we can eliminate by powering vehicles with non-emitting or lower-emitting technologies.
  • In the US, each person uses an average of ~14,000 kWh of electricity per year.
When evaluating alternative energy technologies, we need to think about several key metrics and consequences:
  • Resource availability: How much raw energy is available in a given energy resource? A renewable technology that’s extremely efficient and always available may have very little impact on climate change or pollution if the available potential energy is not enough to satisfy a major fraction of total energy demand.
  • Total CO2 emissions: Emissions can be measured in grams of CO2 equivalent per kWh (CO2e/kWh): In the context of emissions (see clarification here: http://www.skepticalscience.com/Carbon-dioxide-equivalents.html), this unit refers to the amount of CO2 required to produce a global warming effect equal to whatever is emitted by a particular technology, for each unit of energy produced.
    • Lifecycle emissions: How much CO2 (equivalent) is emitted during the production, operation, and disposal of a given technology for each unit of energy generated? Higher lifecycle emissions make a technology less attractive for mitigating climate change and air pollution.
      • The energy payback time is the amount of time for which a power plant must operate to produce the same amount of energy that was required to build it.
      • A longer technology lifetime amortizes the fixed CO2 emissions over a longer period of energy generation, reducing the emission per unit of energy produced.
    • Opportunity-cost emissions: If a renewable generation source takes a long time to plan and deploy, significant emissions may occur during that period; those emissions could have been eliminated if a more quickly deployable renewable source were used instead.
      • The planning-to-operation time encompasses the time to site, finance, permit, insure, construct, license, and connect the technology to the grid.
      • Such delays allow existing, higher-emitting power generation plants to operate longer, and limit our ability to respond to changing energy needs and the need for immediate action against climate change.
      • To minimize opportunity-cost emissions, we want generation sources with a long lifetime and fast planning and installation.
  • Land use and effects on wildlife: Renewable energy sources tend to require more land area than conventional power plants, due to the low energy density of renewable resources. Appropriation of land for energy generation may encroach on or damage wildlife habitats, as well as reducing the total amount of land available for alternative human use.
  • Vulnerability to disruption: Disruptions in the energy supply affect all human activities that are tied to the electrical grid. Centralized energy sources (e.g., nuclear plants) increase the risk of disruption compared to distributed sources (e.g., wind turbines, solar cells), since a single localized catastrophe can wipe out a large portion of local generation capacity.
  • Intermittency: Many renewable resources are by nature dependent on the caprices of complex and unpredictable environmental systems. Wind turbines produce energy only when and where the wind blows. Solar cells and solar thermal converters operate efficiently only when the sun shines on them directly, with no clouds, haze, or human obstructions in between. An energy technology that suffers from high intermittency may not be able to reach high penetration—i.e., to satisfy a large proportion of electricity needs—because it would be unable to fulfill electricity demand consistently. Other technologies must then compensate for the shortfall.
    • Intermittency makes grid-planning more difficult and increases the stress (from ramping) on other generation technologies.
    • The capacity factor is the average fraction of the peak rated power output of an energy source that is sustained continuously over the source lifetime.
      • The capacity factor can be calculated by dividing the total energy actually produced in a given time period by the total energy produced if the source produced 100% of its peak output for that entire period: If a 100Wp solar cell produces 50W for 6 hours a day and 0W for the remaining 18 hours, the numerator would be 50W*6h = 300Wh, the denominator would be 100W*24h = 2400Wh, and the capacity factor would be 300/2400 = 0.125.
      • High intermittency reduces the capacity factor of a given technology.
    • Several strategies may reduce the impact of intermittency:
      • Average out variations by combining generation sources from wide geographical areas. This requires a significant investment in grid infrastructure for long-distance electrical transmission.
      • Smooth out total generation by combining multiple high-intermittency renewable sources (e.g., solar, wind, wave) and using low-intermittency sources (e.g., hydro, geothermal) to provide for excess demand.
      • Use smart meters to charge and discharge electric vehicles at appropriate times to match electricity demand to supply.
      • Deploy an efficient utility-scale energy storage technology: Charge a battery with a solar panel when the sun is shining, and discharge the battery at night to charge an electric vehicle or iPhone. Such a technology does not exist today.
      • Forecast the availability of renewable sources more precisely (e.g., when the sun will shine and when the wind will blow) and ramp up other power plants to offset any excess demand.

Alternative Energy Technologies

Note on energy units: Although terawatts (TW = 1012 watts (W)) are units of power (energy per unit time), I will use it here to describe the total energy available for each renewable technology. To facilitate comparison with our global time-averaged energy production and demand (~15 TW total, of which ~2 TW is in the form of electricity), I’ve converted the total available energy (given by Jacobson in petawatt-hours (PWh) = 1015 watt-hours (Wh)) to the continuous available power (in W) by dividing by the number of hours in a year (8766 ~= 9000 ~= 104 for order-of-magnitude estimates). I think the lack of consistent and personally relevant units cripples clear discussion and understanding of energy issues (what the hell is a tonne of oil equivalent (toe) anyway?)—I like to stick with two units: watts (or kW, MW, GW, and TW) for power, and watt-hours (primarily kWh) for energy.

Solar energy: Concentrated solar power (CSP) and solar photovoltaics (PVs)
  • Essence of CSP: Concentrate sunlight to heat up a thermal transfer medium and generate steam to run a turbine.
    • Leading CSP technologies: Linear parabolic trough, power tower, parabolic dish (no thermal storage)
    • One primary benefit of CSP is the built-in storage capacity: By heating up and storing a material with sufficient heat capacity (e.g., molten nitrate salts (NaNO3, KNO3), steam, synthetic oil, graphite), we can counteract the intermittency of solar energy.
    • Total available energy: ~1000 TW = ~2/3 that of solar PVs (the area required per MW is ~1/3 larger for CSP than for OVs)
    • Capacity factor: 0.13-0.25 (without storage)
    • Lifetime: ~40 years
    • Energy payback time: 5-7 months
    • Planning-to-operation time: 2-5 years
    • Total emissions: 9-11 g CO2e/kWh
  • Essence of PVs: Use an energetically asymmetric semiconductor device to absorb light and generate electricity.
    • Leading PV technologies: Si (amorphous, polycrystalline, microcrystalline), Thin-fim (CdTe, CIGS, CZTS)
    • Solar cells can be deployed in large farms (usually 10-60 MWp, proposed up to 150 MWp) or on rooftops (90% of installed PV capacity is currently on rooftops; a future estimate for that number is 30%). Even the largest solar farms are smaller (in peak capacity) than average fossil-fuel (~600 MWp) or nuclear plants (~1 GWp), hence reducing the risk for disruption of the energy supply caused by centralization.
    • Total available energy: ~1700 TW (~20% of this is realistically harvestable, given the low insolation at high latitudes and competing land uses)
    • Capacity factor: 0.1-0.2 (depending on location, cloudiness, tilt, efficiency)
    • Lifetime: ~30 years
    • Energy payback time: 1-3 years
    • Planning-to-operation time: 2-5 years
    • Total emissions: 19-59 g CO2e/kWh (depends on insolation and amount of energy used in production)
Wind energy
  • Essence of wind: Convert the kinetic energy of moving air (wind) into rotational energy in a turbine to generate electricity.
    • Wind physics
      • The instantaneous power density of wind (usually 100-1000 W/m2) is proportional to the cube of the instantaneous wind speed. Finding locations with high wind speeds is thus crucial for successful wind power generation.
      • Wind speeds at a given height follow a Weibull distribution, which for wind can be approximated quite accurately by the Rayleigh distribution (a special case of the Weibull with shape parameter k=2).
      • Due to shear forces (the friction of air flowing across rough land), wind speeds increase roughly logarithmically with height: Taller wind turbines are thus more efficient. Modern turbines typically have a tower height of ~80 m (the hub or axis of rotation is at that height).
    • In 2007, 94 GW of peak wind generation capacity was installed globally; actual energy production satisfied ~1% of the ~2 TW of global electricity demand.
    • Wind energy requires the smallest footprint of all renewable technologies, although the total land area required is larger due to the array spacing between turbines needed for efficient operation (i.e., each turbine decreases the wind speed behind it, reducing the power density available to nearby turbines).
      • The area required for each turbine is approximately A = 4D x 7D, where D is the rotor diameter.
    • Total available energy: ~72 TW (~6 TW in US)
      • Average wind speeds of over 7 m/s are required for wind turbines to be economically competitive with other energy technologies. Such speeds exist over ~13% of global land area.
    • Capacity factor: 0.33-0.35
    • Lifetime: 20-30 years
    • Energy payback time: 2-4 months
    • Planning-to-operation time: 2-5 years
    • Total emissions: 3-7 g CO2e/kWh (lowest of all technologies surveyed)
Hydroelectric energy
  • Essence of hydro: Release the gravitational potential energy of water in an elevated reservoir, allowing it to flow through a turbine.
    • Hydroelectric plants are built on rivers, either by damming (to flood large areas of land and create a reservoir with water storage) or by introducing turbines that do not significantly alter the course of water flow.
    • Pumped hydro (pumping water uphill, storing it, and releasing it through a hydroelectric plant when electricity is needed) is the only energy storage technology that has been successfully demonstrated at the utility scale (~100 GW, compared to hundreds of MW for all other storage technologies), with storage efficiencies of 70-90%.
    • Hydropower is the largest installed renewable energy source globally, accounting for 17.4% (~36.5 GW) of total electricity production in 2005. Hydroelectric plants are often used as peaking plants since they can start producing power quickly (15-30s when in spinning-reserve mode) to match demand variations and smooth the intermittency of other generation sources—hydropower thus complements other renewables like solar and wind.
    • Total available energy: ~2 TW (5% has been tapped)
    • Capacity factor: 0.42
    • Lifetime: 50-100 years
    • Energy payback time: ~1 year
    • Planning-to-operation time: 8-16 years
    • Total emissions: 16-61 g CO2e/kWh
Ocean energy: Wave and tidal
  • Essence of wave energy: Convert the kinetic energy of wind-driven waves on the surface of the ocean to mechanical motion in a floating generator.
    • The power of a wave is proportional to the density of water (more dense => larger mass (per unit volume) => more kinetic energy), the group velocity of the wave (which is proportional to the wave period), and the height of the wave squared (the intensity or power of any wave is proportional to the amplitude squared).
    • Total available energy: 480 GW
      • Given that 2% of the world’s ~8×105km of coastline has wave power density greater than 30 kW/m
    • Capacity factor: 0.21-0.25
    • Energy payback time: 1 year
    • Planning-to-operation time: 2-5 years
    • Total emissions: 40-60 g CO2e/kWh
  • Essence of tidal energy: Use undersea turbines to harness the energy of oscillatory undersea currents caused by the gravitational attraction of the moon and the sun.
    • The ocean is continuously transitioning between high tide and low tide, with 4 such transitions (2 in each direction) every day. This predictability (6 hours in each direction) makes tidal energy a potential baseload generation source.
    • Tidal currents must have a speed greater than ~2 m/s for tidal power to be economical—lower than the 7 m/s required for wind energy, since water is more dense than air.
    • Tidal turbines are usually mounted on the sea floor, with the rotors either directly exposed or preceded with a narrowing duct to direct water toward them.
    • Total available energy: ~800 GW (~20 GW practical)
    • Capacity factor: 02.-0.35
    • Energy payback time: 3-5 months
    • Planning-to-operation time: 2-5 years
    • Total emissions: 34-55 g CO2e/kWh
Geothermal energy
  • Essence of geothermal: Drive water deep into the ground, allowing the thermal energy in the Earth’s crust to heat it up, and use the hot water to drive a steam turbine.
    • Geothermal energy arises from radioactive decay and planetary formation.
    • Leading geothermal technologies: Dry steam, flash steam, binary
      • Dry and flash steam systems are used when geothermal reservoir temperatures are 180-370ºC or higher; both require the drilling of two boreholes, one for steam (dry) and/or liquid (flash) flowing upward and one for condensed water flowing downward.
        • The dry steam approach uses (vapor-phase) steam pressure to drive a turbine directly.
        • The flash steam approach (most common today) converts hot liquid water into steam in a low-pressure flash tank, which then drives a turbine.
        • Neither system (currently) condenses greenhouse gases (CO2, NO, SO2, H2S) released from the underground reservoir during extraction, instead releasing them to the atmosphere and exacerbating climate change.
      • Binary cycle systems (~15% of current systems) are used when reservoir temperatures are 120-180°C.
        • Hot water rising from a borehole is used to heat and evaporate a low-boiling point “binary” or “working” organic fluid (e.g., isobutene, isopentane); the resulting vapor is used to drive a turbine.
        • Since they are closed-loop systems, binary cycle plants do not emit any greenhouse gases into the environment.
        • Because they can be operated at intermediate temperatures, binary systems may be the most promising and flexible of the geothermal approaches.
    • Total available energy: ~170 TW (but only ~100 GW at reasonable cost)
    • Capacity factor: 0.73 (good)
    • Lifetime: ~30 years
    • Energy payback time: 10+ years
    • Planning-to-operation time: 3-6 years
    • Total emissions: 16-60 g CO2e/kWh

Nuclear energy

  • Essence of nuclear (fission): Use slow neutrons to split the nuclei of heavy elements (e.g., U-235, Pu-239) into high-energy products (e.g., Kr-92, Ba-141, neutrons, gamma rays), which collide with and boil water to drive a steam turbine.
    • The most common nuclear fuels are uranium-235 and plutonium-239.
      • Uranium is typically stored as small ceramic pellets in metal fuel rods, which are used in a reactor for 3-6 years before being replaced.
      • The radioactive waste products resulting from nuclear fission must be stored in proper containment to isolate it from the biosphere for many thousands to millions of years (i.e., until the radioactive elements have decayed sufficiently).
    • The US has more nuclear power plants than any other country (~25% of the world’s total), followed by France, Japan, and Russia.
    • Total available energy: 0.4-14 TW (lower number for once-through thermal reactors, higher number for light-water and fast-neutron reactors)
    • Capacity factor: 0.81 (really good)
    • Lifetime: ~40 years
    • Energy payback time: 1-2.5 years
    • Planning-to-operation time: 10-19 years
    • Total emissions: 68-180 g CO2e/kWh
Fossil energy with carbon capture and storage (CCS)
  • Essence of CCS: Burn fossil fuels (coal in particular, since it’s abundant) but capture and store the resulting carbon emissions to reduce the impact on climate and air quality.
    • Emitted CO2 can be captured and injected deep underground in geological formations (e.g., saline aquifers, depleted fossil fuel fields, unminable coal seams) or into the deep ocean.
      • Worldwide, geological formations have the potential to store up to ~2000 Gt CO2, while we emit ~30 Gt CO2 every year. Clearly CCS is not a permanent solution.
      • Deep ocean sequestration of CO2 makes the ocean more acidic due to the formation of carbonic acid (H2CO3), and the sequestered CO2 eventually equilibrates with surface levels and is re-emitted into the atmosphere.
        • This option has been dismissed by most credible sources.
      • An alternative CCS approach is to combine CO2 with common metal oxides (e.g., quicklime (CaO), MgO, Na2O) to form solid carbonate minerals (e.g., CaCO3, MgCO3, Na2CO3) that can be easily sequestered. This approach requires a lot of raw oxide material and large energy inputs to speed up the reaction (via high temperatures and pressures).
    • Carbon capture technologies can reduce CO2 emissions by up to 85-90%.
      • But the energy required to capture and store CO2 will significantly reduce the power output (by 10-40%) of a power plant equipped with CCS technology.
      • And non-CO2 pollutants are still emitted, at even higher rates than before (more fuel must be burned to generate the same power output).
    • No major power plant today has successfully implemented carbon capture and storage.
    • Total available energy: ~1 TW for 200 years (~60% of annual electricity production)
      • Based on proven coal reserves
    • Capacity factor: 0.65-0.85
    • Lifetime: 30 years
    • Energy payback time: ??
    • Planning-to-operation time: 6-11 years
    • Total emissions: 310-570 gCO2e/kWh
Bio energy: Corn and cellulosic ethanol (E85)
  • Essence of biofuels: Let plants (or animals) convert solar energy into chemical energy, then burn the plants (or animal dung) or decompose them into liquid ethanol for transportation fuels.
    • Biofuels can be solids, liquids, or gases derived from organic matter: wood, grass, straw, manure, corn, sugarcane, wheat, sugarbeet, or molasses. These materials can be burned directly (e.g., for heating, cooking, and transportation) or converted into liquid ethanol for transportation.
      • The sugars and starches in crops are used fermented by microorganisms to create ethanol. Cellulose from switchgrass, wood, or miscanthus (a fast-growing grass) can also be converted into ethanol through fermentation, although the process is far slower than for the crops mentioned previously.
      • E85 fuel consists of 85% ethanol and 15% gasoline.
    • The amount of land required for bioethanol production is massive: We would need to triple the US land area dedicated to corn farming (to ~10% of the 50 states) just to power all US vehicles with corn ethanol.
      • The demands on freshwater for irrigation would be similarly immense.

The conclusion?

Wind and solar energy, coupled with baseload hydro and a bit of geothermal, can turn the climate change equation upside down, slashing emissions while improving the long-term sustainability of our energy supply. Biofuels and CCS do more harm than good. Existing technologies can make a big difference already; technology advances will make renewable technologies more and more attractive by both economic and climate-change metrics.

So where do we go from here?

Start reading. Start learning. Start protesting.

Thanks for reading!

And thanks to my ONE Labmates—in particular Patrick and Geoffrey—for many frightful discussions on energy and climate change.


E.L. Doctorow on writing a novel

July 8, 2013

Writing a novel is like driving a car at night. You can see only as far as your headlights, but you can make the whole trip that way.

I think grad school works the same way.


200 miles in pictures: Reach the Beach 2013

June 1, 2013

A couple weeks ago, I ran the Reach the Beach Relay, a 200-mile footrace across Massachusetts. I’d never met anyone on my team before 9AM that Friday morning, when we gathered across from the Stata Center on Vassar St. at MIT to pile into two Dodge Grand Caravans and make our way into the wilderness beyond Greater Boston. A trunk already packed high with Nutri-Grain bars, fruit snacks, and Gatorade was further burdened by my duffel bag packed with running shoes and bananas.

I was enlisted for this drama by my labmate Christina, who knew a team of MIT chemists (“12 Angry Scientists”) looking for a happy engineer to fill out their roster. I had no idea what I was getting into, but signed up on a whim months ago, then promptly forgot about it in favor of working on my masters thesis. The morning of the race, I turned in my completed SM thesis to the EECS department and clambered into Van 2 carefree.

What it takes to feed 12 scientists for a day.

What it takes to feed 12 scientists for a day.

Twelve team members all accounted for, we headed to Wachusett Ski Resort in central Massachusetts, where we were subjected to a team picture and a quick orientation (“Wear reflective gear at night. Don’t stop in the middle of the road. Recycle. Drink beer—but not too much. Please do not answer ‘nature’s call’ on private or town property.”). The race would be broken into 36 legs of up to 9 miles each—3 legs per person—with a slap bracelet serving as a progressively sweatier relay baton to be passed (slapped?) from runner to runner at the transition areas. Each 6-person van would roll through their full lineup—around 4 hours of running—before handing off the baton to the next van, then rinse and repeat. We would run through the night, through forest and farmland, along highways and dirt roads, until we smelled seawater at the Atlantic coast.
Ski resort in winter, relay race start in summer.

Wachusett Mountain: Ski resort in winter, relay race start in summer.

Slap bracelet baton.

Slap bracelet baton.

Van 2 reporting for duty.

Van 2 reporting for duty.

Faster teams are assigned a later start time so that all the teams end at around the same time, making for a more dramatic finish. It turns out angry scientists = fast runners: we were one of the last teams to get going, 5 hours after the first team started their voyage. At 1PM, our lead-off runner Dan lined up at the start line at the base of Mt. Wachusett. The chair lifts are there for a reason—that reason apparently does not apply to runners. Dan’s 2.8-mile leg was a black diamond trail run in reverse, with 1.7 miles up the mountainside and abundant cursing. But back in Van 2, we had 4 hours to kill before our first leg—we cheered on our Van 1 teammates during their runs by blasting high-quality music like Call Me Maybe and Taylor Swift with the windows down, then stopped for a lunch of turkey sandwiches at a roadside convenience store in the middle of nowhere.
RTB starting line.

RTB starting line.

By 5PM, we were getting antsy in Van 2. After getting all pumped up to start the race, repressing the adrenaline for hours took more self-control than I could muster. So I gave up and took a nap:
Pre-run nap at Assumption College in Worcester.

Pre-run nap at Assumption College in Worcester.

8th in the Angry Scientist rotation, I woke up to grab the baton from #7 Jen and run my first leg, 7.53 miles from Worcester to Boylston. I knew I went out way too fast but couldn’t help it—it was one of the first hot days of the year, I’d been sitting in front of a computer writing a thesis for the past month, and a mile in I was panting like an overexcited puppy. Smooth. Luckily there was no one around to see me self-destruct—my teammates helped with some drive-by dance music—and after 52 minutes of contemplative misery, I rolled into transition area 9 under my own power.

Apparently my preparation was lacking; I’ve run a few marathons (26.2 miles) and half-marathons (13.1) over the years, so I was ready for a calm 7.5- or 8-minute-per-mile pace. Although my total distance here (~22 miles) was similar to a marathon, it was split up into three frantic 6-8 mile races, so I felt compelled to run hard from the get-go rather than pacing comfortably and running to finish. Getting used to the pace was the second-hardest part of this race. The hardest was timing: When to eat, when to drink, when to sleep, when to stretch, when to get warmed up, and—most importantly—when to take a seat on a nearby toilet. No joke. Coordinating alimentary intake and inevitable emission over 24 hours of running is an engineering task far beyond my abilities. Our team captain Kit solved the problem in finest MIT fashion: He simply contracted food poisoning, so that everything he ate came right back up—no need to digest. Brilliant.

Slapping on the baton for my first leg.

Slapping on the baton for my first leg.

A botched hand-off. Sorry Andrew!

A botched hand-off. Sorry Andrew!

#9 Andrew booking it down the home stretch.

#9 Andrew booking it down the home stretch.

Van 2: Jen, Kurt, Stephen, and Yifeng, relaxed and ready.

Van 2: Jen, Kurt, Stephen, and Yifeng, relaxed and ready.

The rest of the race passed by in a blur of sleepless zombie running, insane cheering, dance music, and sweat. We ran over the proverbial river and through the woods and all through the night—creeping up in our van and whispering soft nothings at our runners instead of shouting encouragement—with a brief recess at a local hotel.
Stopping for a late-night dinner of spaghetti and meatballs.

Stopping for a late-night dinner of spaghetti and meatballs.

And... back to running, nighttime edition.

And… back to running, nighttime edition.

The next morning, we finally saw a proper hand-off, 160 miles in.

The next morning, 160 miles in, we finally saw a proper hand-off.

Why is Santa Claus here? Not impressed.

Why is Santa Claus here? Not impressed.

The brothers Horning, celebrating something.

The brothers Horning, celebrating their awesome color coordination.

Thanks to our late start and relative speediness, we passed progressively more competing teams as the race dragged on (I passed ~20 people in 3 legs, and was passed once myself, by an old guy who left me in the dust with his relentless uphill pace). Of particular intrigue was a team named GURL Boston All-Stars, composed of all guys. They were a mystery, and they were fast. At the last transition area, after 192 miles and 24 hours of running, the anchor for GURL Boston stepped into the hand-off zone wearing a short leopard print dress and carrying a clutch purse. Turns out GURL = Gay Urban Running League. As we passed their anchor—the guy had to be running sub-6:30 pace—our van started blasting It’s Raining Men (complete coincidence, of course) with the windows open. He pranced and blew kisses and mimed his thanks, hands to his heart, still running at top speed. What a team.
Damn, GURL.

Damn, GURL.

Our van pulled into Horseneck Beach in southern Massachusetts around 1:30PM on Saturday, with our intrepid driver and final runner Yifeng in hot pursuit. 12 Angry Scientists joined in for the last 100 meters of the race, crossing the finish line after more than 25 hours in transit. We were met with medals and Boloco burritos, and drove back to Boston in delirium.

Reach the Beach 2013 exceeded all my expectations for a relay race. Running for and with a team is infinitely more fun than running alone: you can cover a lot more distance, you get to travel with a built-in fan base, and there’s always someone around to feed you. This particular race was an incredible opportunity to get out of Boston and explore the New England countryside in all its glory: I’ve now used Porta-Potties all across Massachusetts. Who’s in for next year?

The beach, reached.

The beach, reached.

The whole team at the finish: Scientists do it better.

The whole team at the finish: Scientists do it better.

RTB 2013 by the numbers:
  • PB&J sandwiches consumed: 4
  • Nutri-Grain bars consumed: 5
  • Bottles of water emptied: 11
  • Changes of clothes: 4
  • Number of Porta-Potties visited: 9
  • Hours of sleep: 2.5
  • Number of Taylor Swift songs: Too many to count
  • Total distance covered: 200 miles (22.47 for me)
  • Total time: 25:07:39 (2:39:12)
  • Average mile pace: 7:33 (7:05)
  • Ranking: 11th of 147 overall, 4th of 15 in Mens Open division

The photos in this post were taken by me, Andrew, Jen, and Monica. Thanks!


Energy-critical elements: Why you should care about chemistry

January 20, 2013

Image

I went to a talk earlier this week by Robert Jaffe, an MIT physics professor. Jaffe does research on particle physics and quantum field theory, but this talk was on energy-critical elements (ECEs)—chemical elements that (a) are critical for one or more new energy-related technologies and (b) don’t currently have established markets and hence might not be available in large quantities in the future.

For renewable energy technologies, materials availability is particularly crucial because renewable resources tend to be diffuse. Consider solar energy: In a typically-sunny place like the Bay Area, light from the sun reaches the Earth with an average power density of ~200W per square meter. With a record-efficiency 1-square-meter solar panel (and high-efficiency storage), you might be able to power one incandescent light bulb around the clock. Not exactly awe-inspiring. (I recommend LED lighting anyway.) The amount of power we can get from a renewable resource is proportional to the area we can cover with solar panels or wind turbines or tidal energy harvesters, which in turn is proportional to the amount of material we need. To scale up renewable generation is to scale up production of the materials used in renewable generation technology—hence the importance of energy-critical elements.

Here are a few examples of current energy-critical elements and why we care about them:

(1) Tellurium (Te) - Used in advanced solar photovoltaics
    • Tellurium constitutes ~0.0000001% (one part per billion) of the Earth’s crust—it’s rarer than both gold and platinum.
    • Cadmium telluride (CdTe) currently dominates the thin-film solar cell industry (see First Solar). Suppose that this year we want to produce enough CdTe solar cells to generate 1 gigawatt-peak (GWp) of electricity—enough to power on the order of 500,000 homes (less than 1% of the US). We’ll need ~80 tons of tellurium, around 20% of global tellurium production. To get even 50GWp from CdTe solar cells, we need to increase tellurium production by an order of magnitude. As a point of reference, the world consumes around 2 TW of electrical power (average).
    • Conclusion: Tellurium-based solar cells alone won’t solve the energy problem.

(2) Neodymium (Nd) and praseodymium (Pr) - Used in wind turbines

    • These rare-earth elements (REEs) are powerful permanent magnets used in wind turbines. Offshore wind turbines—which are often difficult to repair, for obvious reasons—benefit most from the use of reliable permanent magnets rather than more-complex induction-driven electromagnets.

(3) Terbium (Tb) and europium (Eu) - Used for lighting and displays

    • These rare-earth metals form oxides that serve as red (Eu2O3) and green (Tb2O3) light-emitting phosphors. Such phosphors are used in color TV displays to form subpixels and in standard white lights (e.g., CFLs) to approximate the warm color of incandescent bulbs.
    • The price of Tb and Eu has fluctuated in recent years due to China’s restrictions on exports of rare-earth elements.

(4) Rhenium (Re) - Used in advanced high-performance gas turbines

    • 70% of the world’s rhenium production is used in jet engines (alloyed with nickel) to reduce mechanical deformation under high thermal and structural stresses.

(5) Helium (He) - Used in cryogenics and many research applications

    • Helium is pretty cool: It liquifies at a lower temperature than any other element and doesn’t solidify at any temperature. It’s also the least chemically active element and the only element that can’t be made radioactive when exposed to radiation.
    • Helium is used mostly for energy research, cryogenics—including for MRI scanners—and as a purge gas to flush out liquid hydrogen and liquid oxygen fuel tanks for rockets, since it doesn’t solidify even at ridiculously low temperatures.
    • But we’re running low: Helium is a byproduct of natural gas production, but since it’s lighter than air, it tends to float off into space once it’s released. And we can’t realistically recover helium from the atmosphere, where it resides at concentrations of ~5ppm.
    • We need to efficiently collect the helium released as we extract natural gas; otherwise we’ll quickly run out of a critical element with no known substitutes (at least until nuclear fusion becomes feasible, i.e., perpetually 50 years from now).

(6) Indium (In) - Used as a transparent conductor in touchscreens, TV displays, and modern thin-film solar cells

(7) Lithium (Li) and lanthanum (La) - Used in high-performance batteries

(8) Platinum (Pt), palladium (Pd), and other platinum group elements - Used as catalysts in fuel cells for transportation

Keep in mind that the “energy-critical” label doesn’t reflect any fundamental difference in the nature of elements within and without this classification, and the subset of elements designated as ECEs will change as energy technologies and our knowledge of materials availability evolve.

Side note: Many “rare-earth elements” are energy-critical elements as well, but are actually more common in the Earth’s crust than most of the above elements; REEs are “rare” simply because they don’t exist as free elements or in concentrated ores that are easily accessible.

Side note 2: Some unlisted elements (e.g., Cu, Al, Si) are also critical to energy technologies, but they have developed markets, exist around the world—i.e., geopolitical issues have less impact on their supply—and are used in many other applications, such that substitutes could be found for those applications and additional supply made available for energy applications if necessary.

Energy-critical elements might not be available because they just aren’t very abundant in the Earth’s crust—the only part of our planet we can reach right now. The crust is mostly made of oxygen, silicon, and aluminum; all other elements exist in very low concentrations and are often hard to isolate and extract. That said, the absolute availability of elemental resources probably shouldn’t be our primary concern. A more insidious barrier to the development of new energy technologies is short-term disruption in the supply of ECEs. Supply volatility causes prices to fluctuate, which in turn disrupts long-term extraction efforts and hinders large-scale deployment of technologies that depend on ECEs.

What constraints could disrupt ECE supply?

One primary peril is geopolitics. When ECE production is concentrated in only a few places in the world, international politics and trade restrictions may dictate the market, which is bad. Take platinum: The vast majority of global reserves of platinum-group metals are concentrated in the Bushveld Complex of South Africa. Technical, social, and political instabilities in South Africa could thus disrupt the availability of platinum, palladium, and other critical elements. Another example is China’s 2010 decision to restrict exports of rare-earth elements, a market in which China enjoys a near-monopoly thanks to its natural geological advantages. Rare-earth element prices spiked briefly before the market readjusted to the current and future possibility of limited supply.

But despite all the political talk about energy independence, keep in mind that the US currently imports over 90% of the energy-critical elements it consumes, and that’s not a bad thing. Different regions have different comparative advantages in the production of ECEs, and only through trade are efficient markets achieved. Complete ECE independence is simply not possible—e.g., we have no viable source of platinum—and even partial independence is not possible without sacrificing many modern technologies. Consider food markets: Can you imagine life in a food-independent US? We can’t grow nearly enough bananas, mangoes, cashews, coffee, or cacao to satisfy our massive national appetite, nor can we survive without. Why then do we expect full ECE independence?

Another potential risk for disruption lies in the joint production of energy-critical elements—particularly In, Ga, and Te—with conventional ores. Nearly all ECEs are extracted as byproducts of the mining and refining of major metals (e.g., Ni, Fe, Cu) with much higher production volumes and more established markets. The problem then is that the demand for ECEs does not drive production: Their availability is thus constrained by how much of the ECE is contained in the ore of the primary product, and supply is dictated by economic decisions based on the primary product rather than the ECE. This lack of market control renders ECE prices subject to the whims and fancies of major metal markets.

Adding to the uncertainty in ECE availability is the artificially low prices made possible by joint production: Since ECEs piggyback on the mining infrastructure already in place for major metal production, their prices don’t reflect many of the fixed costs of mining and refining. ECE prices will remain artificially low until by-production saturates—i.e., when enough demand exists that byproduct production can’t keep up, making independent mining of the ECE profitable. At that inflection point, however, new energy technologies developed and assessed using current ECE prices may not be able to afford the much-higher true price and will thus fail. One current example is tellurium, which currently costs around $150 per kilogram and exists with ~1ppb abundance in the crust. For comparison, consider platinum, which costs around $150,000 per kilogram despite its ~4ppb crustal abundance. Why is tellurium so cheap? It turns out tellurium is a byproduct of the electrolytic refining of copper, and the large market for copper keeps the supply of the tellurium byproduct sufficient to meet current global demand.

Other risk factors for ECE availability include environmental and social concerns—the refinement of ECEs (e.g., rare-earths) is often a highly destructive process involving unpleasant chemicals, which could make ECE availability subject to environmental policy—and long response times for extraction—it typically takes 5-15 years to bring new mines online, which may be too slow to keep up with the deployment of novel energy technologies.

So what can we do about it?

Large-scale coordination by the government is needed to attack so complex a problem as energy-critical element availability. Providing reliable and up-to-date information on the availability of ECEs to researchers and investors will go a long way toward improving the current situation: With sufficient information, we can shift research efforts toward energy technologies with ECE needs that coincide with ECE availability.

Another potential response is to increase efforts to recycle ECEs. Recycling all ECE-containing products could reduce our dependence on new resources. Consider cell phones: Modern mobile devices contain 40 or more chemical elements—the majority of known radioactive-stable elements—and most end up in the back of desk drawers at the end of each 2-year contract cycle. But recycling isn’t a feasible option when considering any growth in the market size, much less exponential growth. Assuming the same efficiency of use over time—e.g., the same amount of Te will be needed to produce a CdTe solar cell with a fixed power output now and 20 years from now—recycling can never keep up with increasing material demands, even with 100% recycling efficiency.

The take-home message: Many new energy technologies rely heavily on a subset of chemical elements (e.g., He, Li, Te, rare-earths). These “energy-critical elements” (ECEs) are not currently produced in large quantities, and thus their future availability is highly unpredictable and dependent on complex economic, environmental, and geopolitical factors. A shortage of these elements could inhibit the large-scale deployment of promising solutions to the world’s energy needs. We need more people and more money dedicated to identifying potential substitutes, informing researchers and the public about ECE issues, and improving the efficiency with which we extract, use, and reclaim these elements.

Check out the full APS/MRS report if you’re interested in finding out more!

 Image

Me and my LED light bulb.

Me and my LED light bulb.


2012 in review

December 30, 2012

The WordPress.com stats helper monkeys prepared a 2012 annual report for this blog.

Here’s an excerpt:

4,329 films were submitted to the 2012 Cannes Film Festival. This blog had 16,000 views in 2012. If each view were a film, this blog would power 4 Film Festivals

Click here to see the complete report.


Follow

Get every new post delivered to your Inbox.

Join 1,497 other followers