The problem with concrete

Concrete is 8% of global CO2 emissions. We make 5 billion tons of it each yr. How can we fix this problem?

First of all, why does concrete produce so much CO2? Most of it comes down to the raw chemistry: CaCO3 -> CO2 + CaO. The process literally releases CO2 as a by-product. 60% of the emissions are from this, and 40% are from the burning of fossil fuels needed to run that reaction.

So there are only three ways to fix the problem. Either you (a) reduce the CO2 emissions from the current process (b) find a way to make concrete using a different chemical process (c) use a new material for building (with the same benefits and price as concrete).

There's a bunch of companies (eg Carbon Engineering) working on (a). Cost is still way above $100/ton. Makes it hard for cement companies to do w/out government incentives. Newer separation tech, or converting CO2 to high-value products (like carbon nanotubes), could help.

Regarding (b), I'm not aware of any different chemical processes to make Portland cement. MIT published an electrochemical cell to do it last yr, but these things don't scale (need a GW energy to make MT of concrete). See http://bit.ly/333gIEP - a spreadsheet I made exploring the numbers on this.

So that leaves you with (c) - different materials to replace cement. Neglecting the difficulty of editing building codes, you'd need to replicate the core properties (strength; 10-80 MPa, soundness, consistency, setting time, etc.) at the SAME COST as Portland cement.

One popular approach to doing this is called geopolymer cement. Basically, you use fly ash (waste product, we make 750m tons of it each year) + activator (such as NaOH) + sodium silicates. https://doi.org/10.1016/j.conbuildmat.2013.01.023 has a great overview. However, they calculate that geopolymer emissions are only 9% less than Portland emissions!

I actually disagree with this. I think if you made sodium hydroxide and sodium silicates using solar power + storage (as opposed to buying them off the market, which uses natural gas), you'd only have 90kg CO2/1 m^3 cement, as opposed to the paper's claim of 320 kg CO2/1 m^3.

I also think this approach would also be cost efficient. Note that LafargeHolcim failed in selling this because they bought NaOH off the market (http://bloomberg.com/news/articles/2019-06-23/green-cement-struggles-to-expand-market-as-pollution-focus-grows) - using solar should lower energy costs quite a bit. This is like the Tesla approach of building everything in-house.

In particular, assuming 3.1 MJ per 1000 kg of sodium silicates, and the need for 103 kg sodium silicates / 1m^3 concrete, and $0.02 / kWh of solar, then you get only $6-8 contribution from sodium silicates as a theoretical minimum, not $116 if you buy off the current market. Please let me know if I'm making a mistake here!

The major downside to this approach is that using fly ash ignores the negative externality of coal. Even though fly ash is a waste product from coal combustion, this approach still technically relies on coal power plants to work. Therefore, it's not a viable long-term solution. You wouldn't want to run a company where you're sad that coal plants are being shut down.

Really curious what y'all add to this! Please get in touch if you'd like to think about this w me.

Request for startups #1

Bacteriophage therapy

Antibiotics suck - they wind up creating resistant bacteria, and they're expensive to manufacture. The CDC has declared microbial resistance a threat to global health, predicting 10 million deaths by 2050. Phage therapy would instead use viruses that bind selectively to the cell receptors of bacteria to cause the bacteria to lyse. We've recently isolated phage-encoded lytic enzymes so there's a lot we could potentially do here. 

Energy-efficient HVAC

Buildings are one of the largest consumers of world-wide energy, and we're still far from Carnot's theoretical limit when it comes to building efficient cooling systems. Geothermal and air-source heat pumps have fairly good CoP's, and you could get farther by separating out the humidity system from the cooling system. Billions of ACs will be sold in the next few decades (especially in India) so this goes a long way towards mitigating climate change.

Adoption

Many families in the US wait > 2 years to find a child to adopt. There are many fragmented adoption agencies, each with connections to local churches etc. that charge $40k-$50k per adoption (mostly for legal services). One could perhaps build a SaaS solution to help families find the best adoption plan (it's legal to have an adoption agency in State X, an attorney in State Y, etc.) and minimize legal costs through automation. Perhaps you could even build an adoption agency licensed in all 50 states that provides a really streamlined (but diligent & thoughtful) process for the whole thing. 

Nuclear ships

90% of global transport is through container ships, and these typically take 2 weeks to cross the Atlantic. That's because they actually go slow (10-15 knots) to save money on oil (drag ~ v^2). As modular nuclear reactors (SMRs) become safer and standardized, one could in theory build nuclear ships that would be better for the environment and deliver packages faster. How would that impact the global supply chain?

Feel free to ping me if you're working on any of these or would like to chat about them! My email is first_name dot last_name at gmail.com.

Networking in Silicon Valley

1) To meet founders - become an early adopter of products you like. Get good at something important (consumer marketing, infrastructure, fundraising, etc.)

2) To meet VCs - share your thoughts on a space, find ways to help their portfolio companies with hiring.

3) Become the default person for X. Ex. Ben Thompson for tech strategy.

4) Find the aggregators in your network. Examples: founders of Product Hunt/LunchClub, conference organizers, podcast operators. They're often more approachable than you think. 

5) "Who's the smartest person you know who I should meet?"

6) Everyone's going through a struggle you don't know about. Love people to death.