,
15 tweets,
4 min read
Read on Twitter
PJM's 10-hour duration requirement doesn't pass the sniff test when you look at how the system actually operates on the most challenging peak days - a thread
Some background: PJM has proposed that a resource can only get paid from the capacity market if it can inject over 10 continuous hours. So if you build a 100 MW energy storage system w/ 4 hour duration, your check will be 40% of the check sent to the owner of a 100 MW gas peaker.
PJM is an outlier. Every other ISO/RTO with a capacity market gives full, or nearly full, credit if you can dispatch for 4 hours. But PJM has argued that their peaks are 10 hours long, and so they need 10 hour dispatches to maintain a reliable system.
A reasonable person would conclude: PJM's peak shape must be "flatter" than others', that almost all generators are called on to operate for 10 continuous hours during the worst peak days, and that displacing these generators w/ 4 hr storage would endanger system reliability
By the way, technically the relevant question isn't how long a generator operates on a given day, but how much energy it injected vs. its capacity. If a generator's (total energy injected / capacity) <= 4, then a 4hr storage system could have done the same job that day
So, how does it actually look on some of the toughest peak days in PJM's history? We ran the numbers using EPA's CEMS emissions data from all thermal units in PJM's territory, using the generator's net summer capacity
Let's look at the infamous Polar Vortex Day of 2014 (which also was the longest PJM performance period for at least the last 8 years). 20.4 *gigawatts* of thermal plants ran less than or equal to 4 hours, and the system remained stable. Bulk storage installed today for scale. 
This means 20.4 gigawatts of 4hr energy storage could have replaced these plants on polar vortex day with no impact to system reliability. But under PJM's rule, the energy storage owners' capacity payment would have only been 40% as large.
How about the all-time PJM winter peak day (2/20/15)? 22.4 GWs of thermal plants ran 4 hours or fewer and system reliability was maintained
How about PJM's 2018 winter peak (1/15/18)? 18.7 GWs of thermal generators ran 4 hours or fewer and system reliability was maintained
How about PJM's 2018 summer peak (8/28/18)? 20.8 GWs of thermal generators ran 4 hours or fewer and system reliability was maintained
I'm a storage guy and in all honesty I was expecting the numbers to be maybe 10-20% of what they actually are. Even if the capacity rule is changed to allow 4 hour storage to get fully paid, I would bet we're at least 10 years away from seeing 20 GWs of energy storage in PJM.
Why should we care? By excluding 4 hour storage from competing in the capacity market, supply is artificially low, capacity prices (and emissions) are artificially high, and customers pay higher prices than needed to keep gigawatts of low-utilization fossil fuel generators online
In sum, it seems that PJM's "10 hour rule" does not reflect what the system has actually needed on its most challenging days, and that a "4 hour rule" would be sufficient for a long time to come
I know there are a lot of people at @pjminterconnect, @storage_ESA, and #energytwitter who have thought deeply about this. What data / insights are missing?
