Today, Chilean mining giant
Antofagasta Minerals announced that it had sold its minority interest in the
Javiera photovoltaic solar power plant (though it will continue to off-take
electric energy from it). Looking at the
photo provided with the press release, one can’t help reflect on the challenges
facing renewable energy sources, solar in this case.
The most obvious is that solar takes a lot of space!
The Javiera plant shown here
has an installed capacity of 69.5 MW, yet it covers hundreds of acres.
The largest solar plant to
date is in Kamuthi, India; its capacity is 648 MW and it reportedly covers a
very tight 10 sq. km (2,470 acres) for a land to capacity ratio of 3.8 acre/MW. The most efficient large scale one in the US,
the 550 MW Topaz Solar Farm, occupies 25 sq. km for a 11.2 ratio.
Concentrated Solar Power
(CSP) plants use parabolic mirrors to heat a fluid which then generates steam to
drive traditional turbines. They are
more efficient than the photovoltaic types.
Yet even in the most favorable environments they still occupy a lot of
ground: Mojave One with a nameplate
capacity of 280 MW covers1,765 acres for a 6.3 ratio. The ratio is 6.8 at Solana, 15 at Crescent
Dunes and 4.5 at SEGS.
In comparison, a 56 MW GE
LM6000 combined cycle gas-fired power plant has a much smaller footprint than
Javiera, as evidenced by the photo on the left.
Nuclear plants too are far
more space efficient. And they generate
a lot of electricity. The Millstone
Nuclear Plant, CT, has a capacity of 2,037 MW on a 500 acre site for a ratio of
0.25 acre/MW. The ratio is 0.16 at Peach
Bottom, PA, and 0.4 at Susquehanna, PA.
Clearly, the desert of
Northern Chile is ideal to find ample space for a solar farm, but not all lands
are so devoid of potential use.
The second great challenge is
capacity utilization. In the US, nuclear
plants produce energy more than 92% of the time. By contrast, solar photovoltaics average 18%
and CSP types can exceed 30%. Javiera is
probably in the mid 20s ballpark given the extremely favorable climate of
Northern Chile. But how many places in
the world offer ideal climatic conditions and a ready customer (in the case of
Javiera, a copper mining concern)? Instead, generators in desolated places will
need to build hundreds of miles of transmission lines to connect with populated
areas.
Finally, there is the cost
factor. A popular methodology is the “levelized
cost of electricity” or LCOE whereby initial capital costs, operating costs and
fuel costs are combined over the useful life of a generating plant to calculate
a lifetime cost of generating a MWh.
In the US, the EIA has the
following LCOE estimates for plants entering service in 2020:
(in US$/MWh)
|
Minimum
|
Average
|
Maximum
|
Solar- photovoltaic
|
97.8
|
125.3
|
193.3
|
Solar- CSP
|
174.4
|
239.7
|
382.5
|
Advanced Nuclear
|
91.8
|
95.2
|
101.0
|
Coal-conventional
|
87.1
|
95.1
|
119.0
|
Coal-IGCC
|
106.1
|
115.7
|
136.1
|
Natural gas-advanced
|
68.6
|
72.6
|
81.7
|
It should be noted that these
projections depend on the assumptions made, in particular, the future price of
fossil fuels. Nevertheless, some
patterns are apparent:
1.
Natural gas is by
far the cheapest fuel for electricity generation,
2.
Solar energy is
cheap in places like the Mojave Desert and Arizona because of exceptional
climatic conditions, not so elsewhere,
3.
Coal, natgas and
nuclear plants costs don’t vary much with location.
However, the above LCOEs are imperfect
in two fundamental respects:
1.
They do not
include the cost of standby plants for solar, which is substantial since solar
capacity utilization averages under 20%, and
2.
They assume that
all plants sell their electricity at the same price. In reality, an intermittent producer like
solar doesn’t have the luxury to pick when to sell electricity and could well
be stuck with off-peak demand and prices more often than fossil fuel fired
power plants.
Still, solar technology has
made great progress in the last decade. It
likely will continue to do so. However,
it is crucial for these producers (and the taxpayers) that energy storage
technology advance to improve the economics of the system. Rather than paying out excessive subsidies,
the states and the federal government should invest, or encourage more investment,
in R&D.