PV in California: Looking Deeper

A couple of weeks ago, we discussed a hypothetical case study: the payback period of a 3.7-kW PV system for a PG&E customer in California. One of our readers pointed out that our analysis relied, perhaps too heavily, on a couple of overly optimistic averages, namely that California-based customers could count on 8.2 hours of useful sunlight each day. (Thanks for bringing this to our attention, Jernej!)

This week, I’m going to dig a bit deeper, looking beyond averages to discuss some of the more thorny issues surrounding pricing and panel performance. As it turns out, for the case in question, a pay back period of about 12 to 13 years remains appropriate for customers who use a lot of electricity, and thus pay a high per kWh rate. For customers who pay about 16 cents per kWh on average, the payback period should be in the neighborhood of 18 years. This is, admittedly, 5 years longer than what was originally put forth. But it’s far from the 36 years suggested by Jernej. Let’s take a look.

To begin, PG&E employs a tiered pricing scheme for residential customers, which penalizes big consumers. It looks something like this:

Tier 1 (baseline): 12 cents/kWh

Tier 2 (101 to 130% above baseline): 12 cents/kWh

Tier 3 (131 to 200% above baseline): 21 cents/kWh

Tier 4 (201 to 300% above baseline): 30 cents/kWh

Tier 5 (above 300% above baseline): 35 cents/kWh

As you can see, as a customer uses power above and beyond the usage baseline, the price of electricity increases. Naturally, this creates an incentive to conserve electricity, which is a good thing. Note that winter electricity rates are about 40% cheaper than summer rates. Note, too, that rates will also vary a bit depending on which PG&E schedule you’re on. (For a bit more info, see this previous post). We could literally spend a year’s worth of blog posts discussing electricity pricing, as utilities’ schemes are notoriously complicated. This would, however, be about as interesting as watching dirt. Accordingly, a few observations are relevant:

First, if you commonly find yourself creeping into tier 3 or 4, or even tier 5 territory, your PV system will pay for itself faster than it would for a customer that pays tier 1 or 2 rates. Duh, right? This is because your system will save you more money. In many instances, in fact, a properly designed system can eliminate all upper-tier charges, reducing electric rates by more than half. The second observation is that, in 2008, the average retail rate is closer to 16 cents per kWh, not the 14 cents we used a couple of weeks ago. Finally, electricity rates in California have been increasing by about 5.5% each year, for the past two decades. Our original analysis took 4%, a rate that better reflects national trends. Taking the preceding factors into account, we’ll see that the case for PV is a wee bit stronger than was assumed a few weeks ago.

Now on to our assumptions regarding panel performance. As Jernej pointed out, the assumption of 8.2 hour of “useful” sunlight each day is indeed a bit too rosy. More specifically, the chances that a 3.7-kW system could crank out more than 10,000 kWh over the course of a year are damn slim (at least with today’s technology). This time, instead of assessing expected panel performance on the basis of average sunlight per day, I’ll take an average output rate for PV in the San Francisco area. A conservative figure is 1,580 kWh per year per kilowatt of AC power, as defined by the California Energy Commission (CEC) and performance test conditions (PTC) rating. Put simply, the 187 watts listed on our example Kyocera panels means that, in optimal conditions, the panels will generate 187 watts of power. Under real-world conditions, however, the panels will generate somewhat less (about 167 watts). This second number is reflected in the panels’ PTC rating. The CEC continually conducts studies on a wide range of products, to determine how panel performance is affected by dust, weather, transmission/wiring loss, etc. (For details on CEC and PTC ratings and eligible equipment, see the Go Solar California website).

Details aside, the important point here is to figure out how much power our 3.7-kW system will likely produce. Here we go:

167 PTC watts AC x 20 panels =3340 watts = 3.34 kW

3.34 kW x 1,580 kWh/year/kW = 5,277 kWh/year

Again, thanks to Jernej for flagging our original estimate, which was almost twice as large. Plugging this revised figure into our analysis, we get a rough payback period (without accounting for time-value of money) of 16 years. Alternately, if we assume that the customer pays, on average, 21 cents per kWh, then the payback period drops to 13 years. Take a look at the spreadsheet (xls), and tinker around for yourself.

As we’re all still learning, we welcome questions, comments and suggestions. Next week I’ll continue exploring PV as an investment. Thanks, and good luck.