Thread by @BobKerns

@SVChucko I'll have to see later if I can access the study. A lot just look at present economics, but you really need to look at the installed life, and reasonable projections of future market. Deep solar will change things.

But as it notes, there are other reason for residential.

@SVChucko In particular, to reduce fire risk, we're having grid shutdowns. Reducing the impact of those will be important.

@SVChucko OK, it's not available via JSTOR, but I found a bootleg copy of the paper, and also looked at their data. Unfortunately, that doesn't really answer all my questions, but here are a few observations I think are accurate.

@SVChucko 1) They're looking at near-term, not projecting future grid behavior. This isn't surprising; that would be a very tricky and unreliable projection, involving the difference between green/dirty sources, which vary independently, and how far out of balance they get daily.

@SVChucko 2) None of the utilities in their sample look much like PG&E. I haven't been able to find info on PG&E's marginal emissions. Their average emissions are much lower, due to renewables. So let's assume 0.50 kg/kwh of an open-cycle gas turbine peaker plant.

@SVChucko 3) The key to GHG reduction, is to exploit a difference in the marginal emissions. Let's assume a baseload combined-cycle gas turbine with 0.35 kg/kwh emissions. The difference of 0.15 kg/kwh is what we seek to exploit. If we had coal instead, this gets flipped on its head.

@SVChucko 4) It's really a lot more complex than this, because generation may be dispatched by cost rather than emissions, and because different sources have differing limitations on ramp rates. I'm going to simplify and ignore that here, but I don't want to mislead.

@SVChucko 5) But let's consider some bounding cases first. In a 100% renewable grid, there is no CO2 to eliminate. Also, in a grid with all sources having the same marginal emissions, there's no savings to be had.

@SVChucko 6) The real win is when you have a surplus of renewable at times, and high-carbon at other times. California has occasionally had a surplus of solar that sold for a negative price on the spot market. We still were burning fuel, but it would have cost more to shut down a plant...

@SVChucko 7) ... and buy replacement power while firing up a peaker plant, etc.

But you can see that as the ultimate buying opportunity. But is it really zero marginal emissions? Maybe not because that sold power may be displacing carbon.

I buy green power. Somebody makes that up.

@SVChucko 8) And the makeup power is probably not green. The impact of my purchase is to favor new green capacity, not an immediate win.

@SVChucko 9) But you can see, in the extreme of having an actual surplus of solar in the day, storage allows simple replacement of carbon at night.

But we're presently a long way from having that much installed solar+wind, and will get further away when Diablo Canyon closes.

@SVChucko 10) So for now, we're generally trading high/low carbon sources for any savings, targeting that 0.15 kg/kwh.

What the paper looks at is comparing that, vs going for cost based on present-day time-of-use pricing,

@SVChucko 11) They also consider a third case, where you store any surplus you may produce, to use later to maximize your local usage. To the grid, that looks like conservation at peak solar times. It's a clamped delta, with different timing from cost-based or emissions-based.

@SVChucko 12) Anyway, I would describe their result from a different perspective. The reason cost-based doesn't reduce CO2 is because the pricing doesn't match the carbon. I tend to think that's the problem right there.

@SVChucko 13) But they looked at what amount of carbon tax would be required to compensate. They come up with a fairly high number that suggests it may not be viable. I have some questions around that: is the time of use pricing actualy set appropriately?

@SVChucko 14) But setting that aside, I see three areas to consider.
First, as we move further away from carbon sources, we should hit a period with a larger differential and will need a significant amount of total storage. In that period, savings should increase.

@SVChucko 15) But this also interacts with grid storage, and the general need to smooth the "duck curve". Smoothing the load curve allows for less load on transmission and distribution, with potential cost savings and additional flexibility. Too many scenarios to plausibly analyze there.

@SVChucko 16) But importantly, we will be adding considerable amounts of electric transportation. Local storage is an opportunity to buy power at optimal times (cost OR emissions) while charging at convenience. And vehicle charging can often be done at those optimal times, effectively...

@SVChucko 17) ... adding to the residential storage capacity, and in turn, helping to meet the total storage needs of the grid.

Because with deep solar, we will need several times the storage we have now, which is still mostly pumped hydro. It will likely be mostly batteries.

@SVChucko 18) That storage needs to live somewhere. It can be at the point of generation, at the point of use, and within the transmission and/or distribution system.

Batteries do more than shift the load from one time to another.

@SVChucko 19) They provide voltage and frequency support. They supply "reactive power" helping to keep voltage and current in phase. They provide "cold start" capabilities to power plants. These are all services that have economic value. Solar can also provide some of these services.

@SVChucko 20) To the extent that residential take up some of the needed capacity, it frees up grid storage to handle some of this. And in fact, the current 325+ GW of battery power is used more in this mode than smoothing the duck curve. (Note different scale for 6/21)

@SVChucko @California_ISO 21) It will be very interesting to watch how the economics works out, and how storage will be deployed and utilized. Ultimately, it will be essential to reducing carbon, but this early in the game, the impact is marginal.
/END