McGee Young

Carbon-free energy is the most valuable resource on the planet. Carbon-free energy is the most undervalued resource on the planet.

wcgrid.value has two doors: forecast_value for forward prices and settlement_value for cleared prices. Both are pure functions. The caller supplies typed rows; the engine returns a stacked value breakdown and remembers nothing, open-source future proofing x.com/fiercemcgee/st…

I put the methodology in code and left all running state in grid-platform. wcgrid holds pure pricing functions with no database, network, or disk access. grid-platform owns the rows, API, accounts, and map. x.com/fiercemcgee/st…

This is a fact that needs to be internalized by the DER & environmental community. The value that a battery, smart tstat, ev charger, or heat pump provides to the grid may vary greatly based on their deployment location.

The two-grid dilemma - few know it exists, even fewer have tried to solve it, and no one has. Until now: gridsolver.wattcarbon.com/gridsolver?con…

Once the model is isolated, the next question is what happens inside it: how a device, a location, and a stretch of prices become a dollar figure you can defend line by line. x.com/fiercemcgee/st…

A DER's value follows from how you construct the transmission and distribution grids, then fit them together. The model is the answer, so the methodology needs its own versioned boundary. x.com/fiercemcgee/st…

A distributed energy resource sits on two grids at once. One publishes a price at every node. The other keeps its value inside utility planning. GridSolver began with the problem of valuing both at the same time. x.com/fiercemcgee/st…

@HumanLevelJen my plan is to open-source the library side of the work

@fiercemcgee P.S. Will there be a Github for this?

A few months ago, I started working on the question of what DERs were actually worth to the grid. This question led me down a deep rabbit hole of grid data, most of which couldn't actually answer the questions that I had. I started building my own data layers. A grid digital twin

@xiaowang1984 😂

@fiercemcgee I thought being pedantic was the essence of my account 😉

The grid utilization position that Jigar supports is trying to argue that the Big Dig shouldn't happen in Boston because you have plenty of capacity near Vermont on I-91. Yes the way that metric is measured is similarly irrelevant.

@xiaowang1984 you get a lot of shit on here for being pedantic, but this point deserves more respect. I think the DER industry (me for the most part included) tends to vastly underweight the complexity of turning DERs into grid resources.

Is the argument in the thread that utilization in terms of the broadest definition how many MWh we move through the system per dollar spent should be improved? Sure I can get on board with that. But any attempts to measure broad system wide transmission utilization is a fools errand. You should be looking at this on a constraint by constraint basis because that is the only thing that has meaning. Any number and metric you can make for broad system wide level stuff would be so disconnected from reality that it would serve the old adage "once a measure becomes a target it ceases to be a good measure"

@naywilliams @LeTour too bad it's a race for second place...

Honestly really enjoyed this #Fıfaworldcup a ton 🤘⚽️. I have that bummed feeling like when the Olympics end but luckily still another week of @LeTour 🚵‍♀️. Sports is soo much more than Sport! ❤️

Yeah, not actually. That's impossible to read. Doing this as articles.

OK, going to kick off the "How I built GridSolver" series now. Full disclosure, the AI did the writing, after reviewing the PRs in the code base. Some of it is still awkward and cringy, and I'm trying not to obsess. This is going to roll out in 48 posts, 860 total tweets.

The second, grid-constraint-classifier, ran Dominion's DER dispatch as a live service. Two utilities, two repos, two front ends, both leaning on the same toolkit, wcgrid, to turn grid data into value.

I didn't design that wall up front. I backed into it, from two products that were each trying to model the same two grids and couldn't share the work. The first, pge-flex-explorer, screened congestion corridors across PG&E's Bay Area territory for a settlement pitch.