Cutting Energy Use At Building Portfolio Scale — CleanTech Talk With Carbon Lighthouse’s Brenden Millstein Part 1
Recently I had the opportunity to speak with Brenden Millstein of Carbon Lighthouse about MythBusters, building efficiency, the internet of things, artificial intelligence, and guaranteed HVAC cost savings for building owners. The first part of our conversation is now available. You can listen to the full conversation in the embedded player below. Below that embedded SoundCloud player is a brief summary of the topics covered, but tune into the podcast to follow the full discussion.
If, like me, you enjoy the deep nerdy fun of MythBusters, you may actually have seen Millstein. At one point, Adam and Grant tested the story of Archimedes having invented a defensive harbor weapon using polished shields to create a parabolic mirror of many parts to focus the sun’s rays on attacking ships. They weren’t able to duplicate this, and declared that myth busted, but the story wasn’t over.
Millstein was working at Lawrence Livermore National Labs early in his career, and a co-worker was sure that they could make it work. They had access to great machine shops, set up a rig, ran a few tests and were able to burn a penny and write on wood. The video of their work was sent in to MythBusters, which invited them and another team to do a burn off at the studio. If you missed it, you can watch the episode here, but do come back, as what Millstein has been doing since then is a lot more valuable. And if you want a real death ray, then Heliogen’s machine learning focused mirror sets hit 1,100 degrees Celsius quickly.
Subsequently, Brenden did a stint running a $90 million efficiency program for New York. He realized that there had to be a better way, and with the usual group of young startup obsessives, formed Carbon Lighthouse. Its entire reason for being is to stop climate change.
One of the most interesting things to me from this first part of the discussion was that he and his team have solved the landlord tenant split incentive problem. I’ve run into this issue a couple of times. The first was in discussions with Abid Saifee, VP, Strategic Marketing at the US firm OnSwitch Energy, a discussion I wrote about almost two years ago. OnSwitch does automatic layout of solar panels on commercial building rooftops, but runs directly into the split incentive problem.
The issue is that landlords don’t pay utility charges and tenants don’t do capital upgrades to buildings. That means that the owners of the very large number of buildings that were owned by someone other than the tenant don’t have significant fiscal incentives to upgrade efficiency or HVAC. It’s a barely tapped market for building improvements. My guidance to Saifee was targeted marketing at owner-occupied buildings, and when I was trying to get a heat pump startup off the ground almost two years ago, that was our focus as well.
Millstein and his team didn’t flinch, but went straight after that untapped and essential market. Their solution is based on the same contractual model Bell uses to get phone wiring closets in buildings: they pay the landlord for the privilege out of the savings. Everyone wins. If there are a million of savings, the tenants get 40% of them in the form of reduced utility bills, the landlord gets 40% in the form of a payment from Carbon Lighthouse, and Millstein’s firm gets 20% of the savings.
And the savings are guaranteed. Millstein’s team has been collecting Internet of Things sensor data on buildings for a decade, and now is able to predict with high certainty the savings that they can get.
Prior to March 2020, Carbon Lighthouse sent engineers to customer buildings to install the sensor set, and hired local contractors to do related work. That was obviously a barrier to scaling their offer, but essential in the early years as they honed their data sets, AI, and offerings.
COVID changed all of that. Suddenly they couldn’t send anyone into buildings. This turned out to be a huge advantage, if a painful transition. Now they just send a box or two with 50-60 sensors and instructions on where to put them to the building’s facility services staff. The staff place the instruments instead of Carbon Lighthouse employees and contractors. The big dataset Carbon Lighthouse has collected allows them to start making prescriptive guidance about adjustments to HVAC operational sets, lighting regimes and the like which are implementable either by the facilities staff or contractors the staff procure locally.
This saves a lot of time and money all around, in part because the facility staff are already on site and know their buildings very well, so it’s quicker for them than for outside resources. All of a sudden, Carbon Lighthouse’s biggest barrier to scaling has gone away.
An interesting note on this is the risk of Jevon’s Paradox, which in building efficiency programs means that if efficiency goes up, then demand goes up too. This is a variant of induced demand, a well understood — if not by Elon Musk — aspect of behavioral economics. Millstein and team refer to this as rebound, and it’s not an issue for their market. They target commercial buildings where salaries per square foot are orders of magnitude higher than HVAC costs per square foot, so there is zero incentive to skimp on comfort regardless of efficiency and hence no latent demand, and also a distinct lack of windows that open.
The first half of the conversation was fascinating, and the second half — coming soon — was equally interesting.