|« Help Run on Sun Grow! - UPDATED 2x!!!||Solar Eclipse - A Data Geek's View »|
Say what you will about the California Public Utilities Commission (CPUC), but their recent rulings have been strongly in support of the solar industry and on May 24, they did it again. In a unanimous decision, the CPUC voted to increase the cap on net metering, overriding protests from some utilities and misguided “consumer” advocates.
At issue was the manner by which the statutory cap on net metered installations was to be computed. Under existing state law, California utilities are obligated to accept net metering connections from solar power customers until the installed capacity of the utility’s solar customers equals 5% of the “aggregate customer peak demand." The problem before the Commission was how to define the denominator: does aggregate customer peak demand mean the peak demand that the utility as a whole had seen (as the utilities argued) or was it the sum of the peak demand for each of the utility’s customers?
That turns out to be a very significant difference and the CPUC came down on the side of the solar industry. As a result, the total amount of net metered capacity under the law will now be computed to be 5.2 GW compared to just 2.4 GW under the utility’s interpretation!
Unfortunately, the Commission didn’t stop there. Part of the argument from the utilities was the claim - echoed by TURN - that net metering amounts to a “silent subsidy” from the general rate payer to more affluent customers who can afford solar. In their ruling, the Commission authorized a study to investigate that claim and to quantify the cost-benefits of net metering to the larger rate paying community. Wrote the Commission:
The goal of the study will be to provide the Commission and all interested parties, including the Legislature, with a better understanding of who benefits, and who bears the economic burden, if any, of the NEM program. The report should quantify the costs and benefits of NEM to participants and non-participants and should further disaggregate the results by utility, customer class, and household income groups within the residential class. The study should also seek to gather and present data on the income distribution of residential NEM participants. In order to assess the costs and benefits at various levels of NEM implementation, the above analyses should be conducted using multiple NEM penetration scenarios, including at minimum, the capacity needed to reach the solar photovoltaic (PV) goals of the CSI and the estimated NEM capacity under the five percent cap as defined in this decision. The results of such a study then can be used by the Commission to set future policy for the NEM program, with full awareness of the economic impacts of any policy choices on all classes of ratepayers.
Once the study is complete, the Commission is to promulgate new regulations regarding net metering based on that analysis. All of this is to happen by the end of 2014 - but if new rules aren’t in place by then, net metering will be suspended until such time as they are! Again quoting from the Commission:
We anticipate this temporary suspension in the NEM program, effective January 1, 2015, will remain in place pending the issuance of new rules at the conclusion of a rulemaking proceeding we will commence once the study described above is completed. Of course, if the study can be completed and the new rules are issued prior to December 31, 2014, then the suspension of the program in 2015 will not be necessary. But if the post-study rulemaking remains open and incomplete on January 1, 2015, then under the terms of today’s decision the program will be suspended thereafter, and the utilities will not accept any new NEM applications, until the new rules are issued and take effect.
At the risk of sounding cynical, this sure feels like an opportunity for the utilities to game the system and drag out the rule making if they cannot get the deal that they want. This process will bear close observation!
As an aside, there was one odd detail buried in the decision - the numerator in that 5% computation is supposed to be the installed solar capacity. However, the value used is “CEC-AC Watts” which is computed as the PTC rating of the solar panels used, multiplied by the total number of solar panels, multiplied by the efficiency of the inverter(s) used. So for example, an installation of 20 LG 250 Watt solar panels driving 20 Enphase M215 micro-inverters would have a Nameplate power rating of 20 x 250 =5,000 Watts. The PTC rating of the LG 250, however, is 225.2 Watts and the Enphase M215 is 96% efficient, meaning that the CEC-AC rating would be:
20 x 225.2 x 96% = 4,323.84 Watts
But here’s the curiousity - rebates aren’t paid based on CEC-AC Watts - they are paid on the CSI rating. The CSI rating takes the CEC-AC rating and modifies it based on the actual characteristics of the site - shading, attachment method, azimuth and tilt. This is certainly reasonable as all four of those factors combine to have a significant impact on the energy yield of the system - and that is what an EPBB rebate is supposed to be incentivizing. So why doesn’t the net metering cap use the sum of the CSI ratings as its numerator? And if it did, how big of a difference would it be?
We decided to find out. We took the most recent CSI data set (dated May 30, 2012) and created a pivot table that would report by year and for each of the three investor-owned utilities ("IOUs") subject to CSI, the sum of the CEC-AC and CSI ratings and then give us the ratio. Here are our results combining all three IOUs in each year (in kWs):
These are very interesting numbers - the difference between the CEC-AC rating versus the system-specific-corrections-adjusted, CSI Ratings is very small. Now why is that? The simple answer is that you can actually score higher than 100% when you go from CEC-AC to CSI. How can that be? And how often does that actually happen?
To probe a little deeper we decided to just look at systems installed in SCE territory. Here’s what we found:
Indeed, over the total lifetime of the CSI project, systems installed in SCE territory have averaged 101.7%!
(The numbers drop off in San Diego Gas & Electric to ~98% whereas in PG&E territory the average drops to 96%.)
Still, as a solar installer, we had to admit that these numbers were troubling - we’ve done lots of rebate calculations since 2007 and we really don’t recall seeing design factor scores greater than 100%. An hour of experimentation with the online rebate calculator confirmed that experience - we could get configurations to equal 100%, but not exceed it. So what was the source of these results?
Then it hit us - we were calculating based on EPBB rebates - which makes sense because that is where the CSI rating is reported. However, larger solar systems receive PBI rebates - payments made based on actual performance over five years. What would happen if we distinguished PBI from EPBB rebated systems - would that explain our difference? Indeed it would - check this out:
How about that? From 2006 to 2009, neither rebate method averaged over 100%. But starting in 2010, PBI rebates consistently averaged over 100%, even while EPBB averages remained largely unchanged.
So why the difference? In July of 2009, the rebate calculator was changed. (You can find links to both the current and the old calculator on the CSI website.) It appears that the real difference in how the two calculators work is for PBI rebates - where it allows the design factor to exceed 100%. More importantly, the new calculator takes into account single or dual axis tracking configurations for PBI rebated arrays. Scanning the data, we discovered a site that actually has a recorded design factor of 148%! This is a 1MW system and you can see how a substantial number of comparably sized installations using single or dual axis tracking would really skew the results.
So now we understand where the numbers came from, but we are still somewhat troubled by our results. Are any of these utility-scale systems operating under net metering rules? That seems highly unlikely. And assuming that is the case, shouldn’t they be excluded from the net metering cap computation? In which case, using the CSI rating would add an extra 5% to that numerator, and that would be a good thing! Any CPUC Commissioners or other insiders who might care to educate us further, please do so in the comments.