climate tech (electrochemistry, industrial decarb, synthetic fuels, carbon removal, deep biogeochemistry, solar radiation management)
formerly: ARPA-E, Georgetown, MIT (PhD), UC Berkeley (BSc)
ask me anything: https://jmelville.science/ask/
🇺🇸→🇸🇬
jo melville
Loading...
i recall being pleasantly surprised by the odor of pentanol and hexanol as an undergrad
but this is a silly zeroth order approximation that assumes no active cooling, just blackbody radiation
considering the COP of the HVAC/cooling system (maybe ~3 on a bad day, or as high as 15 in the right climate) and the actual realizable energy savings probably drop to a couple percent, if that
estimated PDRC areal radiative efficiency: ~100 W/m²
est hyperscale data center footprint: 10⁶~10⁷ m²
total cooling potential: 10~100 MW
est total hyperscale DC power draw: 100MW ~ 1 GW
very rough OoM math suggests ~10% energy savings may be achievable
bad comic
the photoenergetic power conversion efficiency of solar is ~20%; corn is ~1%. (20× less efficient)
an EV takes ~0.15 kWh/km; an ICE car uses ~0.75 kWh/km. (5× less efficient)
"99% less land per mile driven" is not hyperbole. it's a pretty solid estimate for average vehicles and energy systems.
🔌💡
Can you identify this country from its electricity grid mix in 2000, 2022, and 2025?
Griddle v.1.2.0 has been updated to use the latest 2024/2025 datasets, because lemme tell you: things have really changed in 3 short years!
Let me know what you think:
jmelville.science/griddle/
🔌💡#energysky
Conservatives hate universities because they are broke outside the club, liberals think it’s gauche to brag, and foreigners self soothe by making “100 top universities for elite reproduction” lists that ignore the fact that, like, ETH Zurich and Oxford have a lower R&D budget than UW-Madison
