Read on Twitter
Energy Twitter has struck a nerv. Apparently a lot of you experts don't know basic terms.
I hope this is a short thread but I'll add to it if I have to.
Tonight I present the first two concepts in this remedial crash course to IRP.
I hope this is a short thread but I'll add to it if I have to.
Tonight I present the first two concepts in this remedial crash course to IRP.
1. Reserve margin is NOT the same as operating reserves!
I've seen multiple articles mess this up, let alone countless tweets.
I've seen multiple articles mess this up, let alone countless tweets.
Reserve Margin is the percent capacity you carry over the forecasted peak load.
This is the number that drives new construction. It's a relatively stable number and determined thru LOLP studies.
13-17% is a sane number. 8% is scary thin (ERCOT) and 30% is super bloated (PJM)
This is the number that drives new construction. It's a relatively stable number and determined thru LOLP studies.
13-17% is a sane number. 8% is scary thin (ERCOT) and 30% is super bloated (PJM)
Operating reserves is how much contingency reserve you carry at any given moment.
This is primarily covered in BAL-002.
Three main components. Non-spinning resources, reg up resources, and balance up resources. (I can cover these more in a 201 tweet if you pass this class).
This is primarily covered in BAL-002.
Three main components. Non-spinning resources, reg up resources, and balance up resources. (I can cover these more in a 201 tweet if you pass this class).
But the important part about operating reserves is they cover all the what-ifs on a minute by minute basis.
Tripped gas plants, fluctuating load, forecast error, renewable variability, etc.
Tripped gas plants, fluctuating load, forecast error, renewable variability, etc.
See how reserve margin and operating reserves are related?
One is total tools in the toolbox, the other is tools you can use right now.
One is total tools in the toolbox, the other is tools you can use right now.
2. LOLP/LOLE, loss of load probability/expectation.
1 in 10 years doesn't mean that once every ten years is acceptable for blackouts.
That would assume a static power grid without changing loads/resources.
1 in 10 years doesn't mean that once every ten years is acceptable for blackouts.
That would assume a static power grid without changing loads/resources.
100 year floods don't happen every 100 years!
Same thing here.
Same thing here.
What it does mean is that you plan your system to have enough capacity to cover the peak hour of every day so that the probability of all your variances stack out to 1/(365*10) = 0.02%.
That 0.02% on the peak hour includes Equivalent Forced Outage Rates of equipment, renewables variability, load forecast error, etc.
And while the stats to come up with this can be tricky, at the end of the day all you're doing is calculating your target reserve margin.
And while the stats to come up with this can be tricky, at the end of the day all you're doing is calculating your target reserve margin.
Lower reserve margin, higher chance of not having enough generation
Don't do maintenance on your gas turbines, higher chance of not having enough generation.
Intermittent resources, higher chance of not having enough generation.
Etc.
Don't do maintenance on your gas turbines, higher chance of not having enough generation.
Intermittent resources, higher chance of not having enough generation.
Etc.
LOLP determines how big your toolbox is.
Generation/imports/DR/DSM/EE are all tools in your reserve margin toolbox.
Then you use your tools to maintain operating reserves to fight the blackouts and keep you ACE under control.
Generation/imports/DR/DSM/EE are all tools in your reserve margin toolbox.
Then you use your tools to maintain operating reserves to fight the blackouts and keep you ACE under control.
/Thread?
Discussion on my weather analogy for 1 in 10. https://twitter.com/FiveThirtyEight/status/1298437108655431680
