Thread by @ScipioAcheronus

Prelim: 100% with you on NJ's nuke subsidy being a good thing, I understand that ACP = Price ceiling, individual RECs, and that Cost Cap = Total $ going into REC compliance market. What I'm saying:
Growing RPS target + Fixed lowish cost cap = Gimmick
Tell me where I'm wrong
1/19

https://twitter.com/JesseJenkins/status/986305361991913472

I'll bring us back to Phil 101, P1 + P2 = C, and you can tell me if I'm wrong on my premises or if my arg is invalid:

P1. Just for simplicity's sake, assume that independent variables like load demand, $ per MWh cost of various types of generation remain flat over time.

P2. A growing RPS target means that the number of MWh RECs required for compliance steadily rise along with your RPS target (NJ example used later)...

P3. Fixed ACP means that as RECs generated (& banked) fall below RECs required for compliance, price of RECs should get closer and closer to the ACP (b/c compliance entity would never pay more than the ACP, but would pay [ACP - $0.01] if rational & if supply is scarce enough).

P4. Similarly, as RECs generated rise above RECs required for compliance, price of RECs should get closer and closer to $0 (b/c Gen would sell a REC for $0.01 if RECs so oversupplied you'd otherwise have to throw it away). Complicated by banking, but general principle holds...

P5. So in normal RPS, unless Class I:
(A) becomes profitable to build w/out subsidies; or
(B) becomes unprofitable to build even if subsidy approaches ACP,
Growing RPS target has generated REC supply growing more or less equal w/ growing target, & value of each REC steadyish.

[Tangent] In the real world, states make their ACP higher than REs currently need for subsidies. So typically there will be a bit more RE generation than the RPS target, utilities bank excess RECs to hedge for future, & REC prices hover below ACP but don't plummet to near $0...

P6. Now let's add a cost cap into the mix. NJ defines its cost cap as "seven percent of the cost to customers of the total number of kilowatt hours sold in the State." So, "Value of all RECs lumped together cannot exceed [total retail $ x 7%]" = Cost Cap.
njleg.state.nj.us/2018/Bills/S25…

P7. This is where things start to get interesting. The math is pretty simple:

[ACP x REC obligation] ≤ Cost Cap 👉 ACP controls individual REC value ceiling, Cap inactive
[ACP x REC obligation] > Cost Cap 👉 Cap controls individual REC value ceiling, ACP inactive

Here's why...

P8. As compliance obligor, once cost cap is < what it would cost to cover all obligations with the ACP, you'll never pay ACP for a REC ever again, EVEN IF total generated RECs fall below what's needed for compliance. Most you're willing to pay is Cap / RECs needed for compliance.

P9. Example: As I said yesterday, 2016 NJ had retail total 75,359,371 MWh. Let's call it 100m MWh in 2030, with 50% RPS, so 50m Class I REC obligation.
ACP = $50, so $2.5B cap based on ACP x REC obligation. OK, on to cost caps...

P10. Let's assume NJ's cost cap isn't eaten up by SRECs, so Class I gets up to 7% of total retail cost. Just to be simple/generous, let's say $150/MWh, for a total retail of $15B, 7% of which is $1.05B. Cap (1.05B) / REC oblg (50m) = $21 is the new real cap on the value of a REC.

P11. Let's assume single compliance obligor for simplicity. If I'm in a situation where REC shortage makes it effectively impossible to meet my REC obligations for under $21/REC (i.e. I'm going to hit the cap), I'll just say screw REC market, I'll just buy ACP's until I cap out.

P12. The only possible wrinkle is a year where I have enough cheap RECs (via banking or Class I generation I own) that I can get close to the obligation, then buy marginal RECs on the market b/w $21-$50, but wind up with my average REC cost <$21. Mostly unsustainable situation...

P13. OK, so since the value of each REC is now basically set by a fixed Cost Cap of all RECs rather than an ACP for each REC, we're in a situation where expanding our RPS target no longer keeps the value of each REC steadyish. Now each expansion just lowers the value of each REC!

P14. So in NJ ex., let's say it's 1/1/2030 and we have our 50% RPS, 7% cost cap, $50 ACP, and effective $21/MWh REC subsidy. We pass a bill that keeps everything the same except we instantly raise the RPS to 100% (100m RECs). Now value of each REC = $10.50. What have we done?!

P15. All we've done is tell the compliance obligor that unless he can meet REC obligations for an average cost of $10.50/REC, he might as well just call up the DPU and say "Hi, I'd like to buy 21m ACP credits, please & thanks, and that'll settle my REC obligations for the year."

P16. We're hurting the RE generation that already exists by making their owners have fewer subsidies with which to pay off their financing cost, and we're lowering the incentive to build more renewables, because the value of RECs just went down. All by raising RPS target.

C. Cost caps are NOT the way to control RPS costs, particularly if goal = RPS target rising incrementally. Control costs via a manageable ACP, so in the event of undersupply the "effective cap" (ACPxREC Obl.) isn't too high. ACP works WITH an RPS growth, cost caps work AGAINST it

@threadreaderapp unroll please