The previous analysis of the California Solar Initiative (CSI) database
summarized the market share of various panel makers. Let's now look at
pricing trends for all-in system costs (i.e., the combined cost of the
panels, racks, inverters, wiring, design services, installation labor,
etc). The graph above shows the median per-Watt pricing for systems of
various sizes for each quarter since the end of year 2007, ignoring any
incentives. Why the median rather than average? Because the average
price could be more sensitive to skewed (erroneous) data inputs. The
per-Watt price for each system was calculated by simply dividing the
total cost of the system by the namepate rating (which is peak DC Watts
under STC, or simply, the number of panels times their rated Watts from
their datasheets). As you can see, the median price for small- and
medium-size systems follows a smoother pattern, while the median price
for large systems is a bit jumpy (and less reliable in determining a
precise trend). The explanation: there are over 2000 small
installations and over 200 medium-size installations while less than 50
large installations, per quarter, on average. The summary of the
dataset (with extra columns, such as average, standard deviation, 25%
to 75% percentiles range, and number of systems) is available here.
The effects of the energy- and the ITC "bubbles" are evident in the
elevated prices in Q3 and Q4 last year. With First Solar modules now
selling below $2 per Watt and c-Si just slightly above, the downward pricing trend will surely continue (a good news for the end customer).
blog comments powered by DisqusNow,
at these prices, grid-connected solar probably does not really make
economic sense yet for the customer in California, on a risk-adjusted
basis, even after considering incentives (as it remains a 20-25 year
investment with various uncertainties about annual AC generation,
unforeseen maintenance, changes in incentive regimes, peak-load retail
price of utility electricity, etc, etc). But, if you have already
decided to go PV, how do you pick an installer? The CSI database could
help answer that question as well. Here is the list of the top 10 installers by MWs and then by number of systems reserved (and not cancelled) since Dec 31st, 2007:
MWs Installer
39 Chevron
26 SunPower
14 SolarCity
14 REC Solar, Inc.
12 SPG Solar, Inc.
8 Team-Solar, Inc.
8 REgrid Power, Inc.
7 Stellar Energy Solutions Inc.
6 Akeena Solar, Inc.
5 Permacity Construction Corp.
# of sys Installer
2223 SolarCity
1482 REC Solar, Inc.
873 Akeena Solar, Inc.
399 REgrid Power, Inc.
334 Borrego Solar Systems, Inc.
296 SPG Solar, Inc.
232 Marin Solar, Inc.
207 Advanced Solar Electric, Inc
196 Acro Electric, Inc.
182 Sungevity, Inc.
As you can tell, it is still a very fragmented market below the top 5-6 players, so there is plenty of choice.
Now,
back to the main focus of this blog: How do Unisolar's system prices
stack up against the median prices over the past 18 months? An easy
question to answer, as there have been a total of 26 (only!)
Unisolar/SIT systems reserved and not cancelled over that period, so we
can plot them all (5 systems are priced over $11 per watt, and thus,
off the chart).
There
were no large (100KW+) Unisolar or SIT systems reserved and not
cancelled over the period, so the myth that somehow Unisolar is the
preferred solution for large commercial rooftop PV is busted. Another
myth, that Unisolar's and SIT's all-in system costs are somehow lower
than competing rooftop PV systems due to the allegedly easy
installation, is busted as well. While for the small systems Unisolar's
pricing is all over vs the median price, for medium-sized systems, each
and every Unisolar (and one red-dotted SIT) system is priced above the
median (and all but one are among the 25% most pricey systems each
quarter). No wonder Unisolar's market share has been just 0.1% over the
past 18 months, according to the CSI database (only 0.3MW reserved and
not cancelled). Oh, and if you are wondering how many of those top-10
installers had a Unisolar installation reserved (and not cancelled) in
the past 18 months, the answer is: none.
Who would miss Unisolar in California, if they are to become ineligible for incentives
in two days? And would ECD's auditors ever bother to look into what
really happened to most of those reported shipments to North America -
the 19MWs in calendar 2008 (plus up to 4MW, or 20% of 20MWs in the March 2009 quarter)?
