82 Billion Dollars

 If we spent 82 billion dollars on PV at $4/watt installed in the Mohave
desert with a grid tie, that would be 20.5 GigaWatts. Panels operate at 75%
of their rating, so 15.4 GW. Lets say we get 5 hours of standard sunlight a
day for 30 years, that would be 8.42*E+11 kW-hrs, so this would be $0.097
$/kW-hr electric power.

2 ways to improve on this are lower cost PV and longer lifetime.

 
Your calculation is too simplistic. Power is not distributed as DC, so the
energy needs to be converted into AC. Your costs also do not include the
cost of the infrastructure to support the panels, or the people that would
be needed to support the plant. Not only that, you can't pluck $82 Billion
out of the cookie jar, so that money would need to be borrowed and paid back
(at a cost).

Just as a matter of interest, what happens for the other 19 hours of the
day? Either there must be alternate power sources ( which would be
relatively idle during the sunlight hours), or the energy must be stored.
Either way, there would need to be a huge additional captial outlay to cover
it.

If someone could use solar power on a massive scale and make it profitable,
it would be being built now.


regards,
Mark

 
They are; but:


No and yes.  There are long-distance DC transmission lines, including
two in California (the "Pacific DC Intertie", which runs along US 395
from LA up to Oregon; and the "Southwest DC Intertie", not sure where
that one is).


Actually, they do -- at $4/watt, 82 billion dollars buys 20.5 GW
(as he said above), so you should be able to get a good "bulk buy"
deal :-) ... and panels are currently just over $3/W even in small
quantities, so this leaves at least $1/W for infrastructure.  This
might not be quite enough but his 5 PSH/day number is too low too
-- Inyokern California, which might be a good location for the
project, gets 7.66 PSH/day (according to the NREL "red book").
Trackers would increase this slightly; whether that is worthwhile
depends on the additional energy versus the cost of trackers.


Not too many people should be required, but some maintenance would.


Yes.  Current borrowing costs range from about 2% to about 10%
depending on the type of bond and its duration.


This is the wrong number.

The panels produce power from before dawn to after dusk, with the
peak occurring in the peak sunlight hours.  The production does
not quite match the peak load curve in California (peak load occurs
in late summer, and around 3 PM instead of closer to noon).


The question is not "what happens the other 19 hours", nor even
"what happens the other 24-7.66 = 16.34 hours", because these are
*peak* sun hours, not actual sun-hours.  So the question is really
"what happens the other ~12 hours" -- and the answer is, the load
is supplied from the remaining generation, which already exists.
California's load runs from about 18 GW at night up to about 40-50
GW during the day.  The panels supply from 0 to ~20 GW of this,
with the PV supply nonexistent at 3 AM when it is not needed (the
18 GW minimum coming from elsewhere) and peaking a little too soon.
Instead of being stored in batteries, it would be stored in the
hydroelectric plants, which would simply run less water volume at
those times.


Maybe.  There are two problems here:

 - it *is* being built now; and
 - 10 cents per kWh still costs too much.

Wind power produces energy at ~7 cents per kWh (including interest
payments on the wind farm), and -- if the tax credit is put back
into effect -- is subsidized at about $.02/kWh, so that its producers
break even or profit at $.05/kWh, or "50 mills" as they say in the
industry.  Wholesale electricity in California generally costs
between 30 and 60 mills, depending on time of day and time of year.
The higher prices occur during the summer peaks, which (rather
nicely) is when the PV panels work best.

"Green" power can also be sold at a premium to various customers:
Utah Power, for instance, was charging an extra $.03/kWh for its
"Blue Skies" program (now reduced to ~$.02).  Some Arizona utilities,
which are one of the places building "massive scale" PV (but nothing
like 20 GW!), charge a premium -- I do not know how much -- for PV
power.  Both these programs have in fact obtained customers, so
the stuff *does* sell.

If we re-run the calculations at $6/watt and 7.66 PSH/day, and use
$75B instead of $82B, we get:

    $75B / ($6/watt) = 12.5 GW

    12.5 GW @ 75% efficiency (MPPT/no-battery = ~82%; throw in
        some reduction for high daytime desert temperatures --
        Inyokern and China Basin armed force base are near
        Death Valley and it does get toasty -- and for transmission
        losses) = 9.4 GW out

    9.4 GW x 7.66 PSH/day = 72 GWh/day

    72 GWh/day x 30 years = 789 TWh

so $75 billion buys about 789 terawatt-hours, at a price of about
95 mills (excluding interest charges).  So at $6/watt it does not
work, even with "free" money, unless we can get a premium of at
least $.04/kWh.  This is a bit high; it looks like the available
premium is closer to $.02-$.03 -- and the available premium will
tend to drop as you take out the "high end" customers (those who
are more willing to pay extra for "green power").

Get the price down to $4/watt, as suggested originally, and the
cost for the 9.4 GW PV system drops from $75B to $50B.  Now the
electricity costs only 40 mills (excluding interest charges) and
is quite competitive.  Unfortunately, interest rates on 30 year
treasury bonds are close to 5% today, which makes the project iffy
again, even if you can promise $4/watt.  But this is why, as I said
above, it *is* being done now, on smaller scales, as pilot / test
projects to see what the costs really are, and what they might be
once the "d'oh!"s are fixed.
--
In-Real-Life: Chris Torek, Wind River Systems
Salt Lake City, UT, USA (40°39.22'N, 111°50.29'W)  +1 801 277 2603
email: forget about it   http://web.torek.net/torek/index.html
Reading email is like searching for food in the garbage, thanks to spammers.

 

To get the MVs required for that transmission line would require just as
much equipment. On the other hand, I think it is reasonable to say MOST
transmission lines are AC, so either way, there is an additonal cost.


Should and will are two different things. To place such an order would
require magnitudes of increase in manufacturing capacity. The company
producing these things would need to obtain a LOT of money to be able to do
this. Unless this is staged over decades, the price will not be lower. If it
is staged, the original premise is invalid, since it will be a gradual
integration which I expect.


Without a tracker, the incident solar radiation is less than 90% of noon
capacity for more than 70% of the day and less than 70% of noon capacity for
more than 50% of the day. In addition, attenuation increase as the elevation
drops because of the increased path through the atmosphere.


How are you intending to keep the dust and stuff off the panels? Not keeping
the panels clean can reduce their capacity. Either you clean them or derate
them. Eitehr way, there is additional cost.


A misleading statment. The deliverable power drops away substantially with
lower angles of incidence. At dusk, the power would be a trivial percentage
of the peak capacity. This peak capacity of the opanel also needs to be
derated for the season.


Then the power rating must be devalued to account for the fact that at 3pm
(no tracker), the solar radiation would be (on average over the year) only
be 70% of the noon peak and it gets rapidly worse as the day goes on. I
don't know where such a farm would be put in the US, but I guess in the
30-40 degree latitude band. At this latitude, there is about a 3 time
variation in solar energy between summer and winter, so there would need to
be adjustment to account for these variations as well.


Actually, that is not "storing" of the solar energy. I say this, becasue
unless there is a lot of excess hydro capacity, extra will need to be
provided (at a cost). When a coal fired plant has to provide additions sets
to account for peak demand, this cost is included in the real cost of the
power. Using hydro stations in a similar capacity would need to be included
in the real cost of the solar power.


profitable,

Where is there solar power on a massive scale?


This is a rather expensive top-up method, especially since it is not at its
peak when demand is at its peak. The total lefe cost of the power will be
much higher per kWH if this is how it is used.


One can sell a smelly cheese to a certain percentage of the population. It
does not follow that one can scale up scales by simply making more smelly
cheese.


I suggst this is inaccurate since it does not take into account derating for
seasonal and diurnal variations. Nor does it take into account that solar
power will be grossly affected by normal weather conditions, albeit for
shorter periods in a desert. In conventional power stations, the power is
not normally all or nothing - one or two sets can go down and not have a
serious impact. Once the clouds come, the whole farm will be severely
degraded for the duration of the storm. I know this to be the case, because
we have a solar powered site in the Australian desert and perhaps once or
twice a year we have storms which virtually stop the solar panels from
producing any significant power. This is ok, since we use batteries, but if
those panels were a primary supplier to the grid, there would be trouble.


I do not belive this is possible at this time.


Unless we use different terminology, pilot/test plants do not constitute
massive scale implementations.

Mark