...The photovoltaic plants in operation in Switzerland show an
average yearly electricity production of 819 kWh per kWp.
(Average yearly production in US is 1800kWh - K.)
...An important yardstick for the assessment of renewable en-
ergy systems is the estimation of the energy and/or environ-
mental pay-back time. This describes the time until environ-
mental impacts from the production of the plant have levelled
out due to avoiding resource uses and/or emissions of a con-
ventional reference system that produces the same amount
of electricity. Here, we consider a modern natural gas-fired,
gas combined cycle power plant as the reference system
(Dones et al. 2004). It is assumed that the use of photo-
voltaic power plants can avoid the installation of such a
facility.... the pay-back time for the indicators
non-renewable and non-renewable plus hydro cumulative
energy demand. This time is between 3 and 6 years for the
different PV plants. This means that the energy use for pro-
ducing the photovoltaic plants is as high as the energy use
for the operation of the gas power plant during 3 to 6 years.
Thus, it is five to ten times shorter than the expected life
time of the photovoltaic power plants."
The widespread deployment of PV in the US would be aimed at eliminating COAL. The article acknowledges the sensitivity of their method to the assumption of the energy source it is displacing .
"...Differences for the situation in other countries,
in comparison to the data modelled for Switzerland, are
mainly due to different solar irradiation. It should be con-
sidered that the inventory may not be valid for wafers and
panels produced outside of Europe, because production tech-
nologies and power mixes for production processes might
not be the same. For the modelling of a specific power plant
or power plant mixes outside of Switzerland, it is advisable
to consider at least the annual yield (kWh/kWp) and if possible
also the actual size of the plant in square metres...."
As for your assertion regarding the environmental consequences of PV, here is the full statement from your source. Note the difference from your cherry picked version:
(continues from above quote - K)
"...This picture changes if emissions are taken into account.
Valuating the impact with the method of ecological scarcity
(Umweltbelastungspunkte - UBP) gives an environmental pay-
back time of about 25 years (Brand et al. 1998, Frischknecht
et al. 2004c), while
the pay-back time evaluated with Eco-
indicator 99 (H,A) is slightly higher than for the energy de-
mand. With human toxicity (Guinée et al. 2001) as an indica-
tor, it
might be even higher than 50 years and thus longer than
the expected life time. The reasons for these large differences
between the selected LCIA methodologies are manifold: in
EI99 (H,A), the consumption of fossil resources is considered
as very important, but not so important in ecological scarcity
and not included in the human toxicity category indicator; the
PV chain produces higher emissions of human toxic species
than the gas chain. The outcome of such a comparison is in-
fluenced by the choice of the reference system, on the one
hand, and by the indicator, on the other. The analysis of the environmental impacts with different
LCIA methods shows that it is quite important to include
process-specific emissions of the production chain into ac-
count.
It is necessary to evaluate all types of environmental
impacts with different (midpoint and endpoint) LCIA meth-
odologies, if photovoltaic power plants are to be compared
with other energy systems. The results of such a comparison
are quite dependent on the choice of LCIA methodology. The approach in the ecoinvent database for dealing with
multi-output processes has several advantages. The alloca-
tion is done in a fully transparent way, because all alloca-
tion factors are reported. The LCIs of all co-products in the
database are fully consistent. If necessary, allocation factors
may be changed and LCI results can be recalculated."
Here is a similar analysis proposal with summary of results conducted in Ann Arbor MI.
http://css.snre.umich.edu/main.php?control=detail_proj&pr_project_id=76It's unfortunate that the chart showing the payback periods against fossil and nuclear isn't able to be posted here.