How much ESS do we need (P2)
TL;DR: Most consumption already occurs during daylight and there are things we can do to move more there which reduces ESS needs.
Still working through the implications, see post
#875 for part 1. We need energy storage for when renewables aren’t generating power, but can we make simple tweaks to meaningfully reduce the amount of storage?
The EPA says the world consumed ~25.3 PWh of electricity in 2021, ~101 PWh
e of energy overall (Musk subtracts out fossil fuels need to refine/transport/find fossil fuels at comes up with
82 PWhe as what's actually needed to replace fossil fuels). How much of that can we move into the daylight?
Let's look at it sector by sector.
Agriculture
You can't change when the cows need to be milked or the heat needed to keep the baby chicks warm at night.
The electricity used by the agriculture sector is accounted for in the electricity sector, so the power consumed
in this sector is direct fossil fuels for heat and machinery. Heating occurs primarily during the night and
machinery usage is primarily during the day. Those of you in the sector can make a better call than me, I'd say 70/30, or ~3 PWhe. | |
Transportation
From the graph to the right, most transportation occurs during the day. But, that doesn't help us
as electric vehicles are recharged while they are at rest, not on the road (except the Aptera). Using
batteries to recharge other batteries incurs additional round-trip penalties.
Incentivizing charging EVs during the day at work (or more likely disincentivizing night-time
charging) could lead to a big reduction in the amount of grid storage that would be needed, this
becomes more practical as charging times decrease (e.g., CATL’s new sodium battery can be
completely recharged in 15 min). But it could greatly reduce the need for storage capacity.
Also, as vehicles become autonomous, they could become self-charging. | Vehicles on the road by hour/day
 |
An enormous part of transportation is travel to & from work. Letting employees work from home, where possible, on
dunkelfaute days would greatly reduce the need for that energy that day. That would greatly reduce the amount grid battery backup needed for periods of low production.
So, just as we have snow days, we might have dunkelfaute days. Not traveling on dunkelfaute days would not only greatly reduce transportation energy needs, but also reduce industrial energy consumption as workers would have the day off.
Using dunkelfaute days allows grid operators to have "breathing" space for when weather conditions are worse than anticipated, and not need to tremendously overscale battery backup.
Industry, Commercial, Residential, & Electricity
As industrial, commercial, and residential are primarily the direct burning of fuels for heat, they'll eventually get
combined into electricity; so let's tackle them together.
As was mentioned previously, the graph to the right is outdated as it doesn't show the electrification of heating,
as such winter curves (blue) will more match summer cures (yellow) except that peak of the curve will occur during
the night when solar isn't available.
Based on that, I doubt there's anything northern climates can do to reduce ESS capacity.
We might be able to alter the workday from 9-5 to 8-4 to shift the curve left more towards solar hours. But most
would probably make use of the daylight time to do things outside because there is light and then retire to do
food prep after the sun went down (I would anyway). So again there would be no real savings. |  |
Industry that requires a lot of overnight power might be required to coordinate with grid operators or have their own energy storage.
FERC & Grid Improvements
As mentioned previously, when the sun doesn't shine here, it does shine over there. The U.S. does not ship power coast to coast. But, there is a interstate system that most (The Texas Interconnection is maintained as a separate grid for political, rather than technical reasons) are connected to. This system can be strengthened at lower costs than adding additional days of ESS. The
FERC serves as an economic regulator to prevent abuses, but again the system could be enhanced to make it better.
VPPs
Virtual Power Plants (
VPP) have been in the pilot stage for years and may soon dominate the ESS market. Tesla for example has become an approved energy provider that utilities can buy power from and has a pilot program to sell excess capacity from customers' batteries (either powerwall or EV) into the grid. The money is passed on to their customers, so when the rates are high enough the battery can discharge to some preset rate and the owner makes money off their investment.
With 1,356,203 Teslas on the road, not including any other car maker, assuming they don't let the battery drain more than 50%, that's 68 GWh of ESS in garages immediately available. With 6,809,322 EVs globally at 50% and 100 kWh battery packs that's 340 GWh.
How Much Storage is Needed?
Master Plan 3 had 240 TWh for the world, but I still haven't seen the white paper to review the assumptions.
Using
The Institute of Physics numbers it would be 227 TWh, but their paper is really only looking at Germany.
If you went the old adage of 3 days, you come up with the general number of 674 TWh.
What's that cost over 30 years?
At today's
industrial cost of LFP of 118 $/kWh, than 240 TWh is $0.32/d per adult. With Sodium at $77/KWh, it's $0.21/d/p and if the iron flow battery (
#840 &
#849) can be believed it's $0.03/p per year.
Remember the 1.4 billion cars that needed to be replaced with EVs? If they all had 100 kWh packs and were hooked up to VPPs with no more than a 10% drain that would be 1.4e9 x 100e3 = 140 TWh of energy storage, but of that, only a small percentage would be available as most won't want to drain their batteries more than 20%; so ~28 TWh of storage. But it shows how important VPPs will be.
dailysceptic.org
