WEBVTT

00:00:00.060 --> 00:00:02.490
The following content is
provided under a Creative

00:00:02.490 --> 00:00:04.010
Commons license.

00:00:04.010 --> 00:00:06.340
Your support will help
MIT OpenCourseWare

00:00:06.340 --> 00:00:10.710
continue to offer high quality
educational resources for free.

00:00:10.710 --> 00:00:13.320
To make a donation or
view additional materials

00:00:13.320 --> 00:00:17.213
from hundreds of MIT courses,
visit MIT OpenCourseWare

00:00:17.213 --> 00:00:17.838
at ocw.mit.edu.

00:00:26.016 --> 00:00:28.640
PROFESSOR: Today we're going to
dive into Cost, Price, Markets,

00:00:28.640 --> 00:00:29.750
and Support Mechanisms.

00:00:29.750 --> 00:00:33.370
The support mechanisms
otherwise known as subsidies.

00:00:33.370 --> 00:00:35.107
So this is Lecture 18.

00:00:35.107 --> 00:00:37.190
We're approaching the end
of our course, actually.

00:00:37.190 --> 00:00:39.273
We have about, we have a
handful of lectures left,

00:00:39.273 --> 00:00:42.440
and then we go
our separate ways.

00:00:42.440 --> 00:00:47.280
This particular lecture will be
followed up not this Thursday

00:00:47.280 --> 00:00:49.350
but the following Thursday.

00:00:49.350 --> 00:00:50.960
We'll have a guest
speaker come in

00:00:50.960 --> 00:00:52.585
and talk about the
cost model that he's

00:00:52.585 --> 00:00:56.160
developed for PV into a
very high level of detail,

00:00:56.160 --> 00:00:57.262
so it'll be a lot of fun.

00:00:57.262 --> 00:00:58.720
And you'll be able
to use that cost

00:00:58.720 --> 00:01:02.600
model to model your own
PV devices, apparatus,

00:01:02.600 --> 00:01:04.680
and so forth.

00:01:04.680 --> 00:01:07.850
Next Tuesday, a week from today,
we'll be touring a PV facility.

00:01:07.850 --> 00:01:10.100
It'll be here on campus to
make it easy for everybody.

00:01:10.100 --> 00:01:11.558
We'll go over to
the student center

00:01:11.558 --> 00:01:14.735
and tour the PV system up on the
roof there, as well the balance

00:01:14.735 --> 00:01:15.710
of system components.

00:01:15.710 --> 00:01:17.400
So we'll be able
to have a close-up

00:01:17.400 --> 00:01:19.470
look at how that works.

00:01:19.470 --> 00:01:22.480
But today, cost, price,
markets, and subsidies.

00:01:22.480 --> 00:01:25.550
We want to talk
about those items

00:01:25.550 --> 00:01:28.300
because, at the
end of the day, PV

00:01:28.300 --> 00:01:31.790
is a product that is competing
against bulk electricity.

00:01:31.790 --> 00:01:34.570
And if we can't compete
against the bulk electricity,

00:01:34.570 --> 00:01:39.110
then our on-grid applications
are going to be rather limited.

00:01:39.110 --> 00:01:42.940
So we want to understand how all
this works and fits together.

00:01:42.940 --> 00:01:46.100
I'll be providing you several
snapshots and several pieces

00:01:46.100 --> 00:01:49.260
of the puzzle with a lot of
discussion back and forth

00:01:49.260 --> 00:01:52.710
over the course of
today's lecture.

00:01:52.710 --> 00:01:56.040
First off, let's dive
into PV cost and price.

00:01:56.040 --> 00:01:59.840
What is the difference
between cost and price?

00:01:59.840 --> 00:02:02.830
They're used interchangeably
in colloquial language,

00:02:02.830 --> 00:02:04.150
but there's a big difference.

00:02:04.150 --> 00:02:04.956
Jessica?

00:02:04.956 --> 00:02:07.247
AUDIENCE: Cost might be what
it costs the manufacturer,

00:02:07.247 --> 00:02:09.204
and price is what it's
sold at [INAUDIBLE].

00:02:09.204 --> 00:02:10.120
PROFESSOR: Absolutely.

00:02:10.120 --> 00:02:14.715
So cost is what it actually
costs to make, to manufacture,

00:02:14.715 --> 00:02:16.840
and price is what people
are willing to pay for it,

00:02:16.840 --> 00:02:19.320
what the market is demanding.

00:02:19.320 --> 00:02:22.580
So sometimes price can be
above costs-- you're hopefully

00:02:22.580 --> 00:02:25.360
in that situation most of the
time-- and sometimes price

00:02:25.360 --> 00:02:27.050
can actually be low cost.

00:02:27.050 --> 00:02:30.284
What is an example of when
price could be below cost?

00:02:30.284 --> 00:02:33.010
AUDIENCE: The Amazon Kindle?

00:02:33.010 --> 00:02:34.290
PROFESSOR: The Amazon Kindle?

00:02:34.290 --> 00:02:35.720
Why would it be doing that?

00:02:35.720 --> 00:02:40.030
AUDIENCE: Because they want
get people to adopt the device

00:02:40.030 --> 00:02:43.000
and then make money on
subscriptions to books

00:02:43.000 --> 00:02:43.990
or [INAUDIBLE].

00:02:43.990 --> 00:02:45.970
PROFESSOR: Hm.

00:02:45.970 --> 00:02:47.230
A loss leader, right?

00:02:47.230 --> 00:02:50.755
Something like, for example,
if you buy your razor handle,

00:02:50.755 --> 00:02:54.140
and that's really cheap, but
then they gouge you on blades.

00:02:54.140 --> 00:02:58.540
Or other examples
include the cheap items

00:02:58.540 --> 00:02:59.540
in the front of a store.

00:02:59.540 --> 00:03:01.400
You walk inside the
store, and then you're

00:03:01.400 --> 00:03:03.608
barraged by all the more
expensive ones right inside.

00:03:03.608 --> 00:03:06.800
So loss leading can be one
example of price below cost.

00:03:06.800 --> 00:03:07.300
Another?

00:03:07.300 --> 00:03:07.955
Other examples?

00:03:11.670 --> 00:03:16.390
What if I start
making a gizmo, and I

00:03:16.390 --> 00:03:18.017
pocket an enormous profit.

00:03:18.017 --> 00:03:20.600
And all of you start watching
me make that gizmo and say, hey,

00:03:20.600 --> 00:03:21.183
I can do that.

00:03:21.183 --> 00:03:22.564
It's pretty simple.

00:03:22.564 --> 00:03:23.980
It doesn't take a
rocket scientist

00:03:23.980 --> 00:03:24.830
to manufacturer that gizmo.

00:03:24.830 --> 00:03:25.560
I can do it too.

00:03:25.560 --> 00:03:28.760
And everybody starts
manufacturing gizmos.

00:03:28.760 --> 00:03:32.660
Pretty soon, we overwhelm the
demand, at least at that given

00:03:32.660 --> 00:03:34.680
price point, and the
price is depressed

00:03:34.680 --> 00:03:35.680
as we enter a price war.

00:03:35.680 --> 00:03:38.690
We enter what is called
an oversupply condition.

00:03:38.690 --> 00:03:41.510
That's another example where
price can fall below cost.

00:03:41.510 --> 00:03:43.740
And another reason
why price can fall

00:03:43.740 --> 00:03:46.599
below cost is simply
the price, or the market

00:03:46.599 --> 00:03:48.890
you're trying to address,
simply won't buy your product

00:03:48.890 --> 00:03:49.960
at that cost.

00:03:49.960 --> 00:03:51.960
And that's the case with
substitution economics.

00:03:51.960 --> 00:03:55.792
If we're competing against
fossil fuel-based electricity,

00:03:55.792 --> 00:03:57.750
let's say, and we want
to compete against that,

00:03:57.750 --> 00:04:00.960
we might not be
able to manufacture

00:04:00.960 --> 00:04:03.810
solar panels cheap enough
to address certain markets.

00:04:03.810 --> 00:04:06.270
For example, Wyoming, which
has $0.05 per kilowatt hour

00:04:06.270 --> 00:04:10.920
electricity due to
cheap fossil fuel.

00:04:10.920 --> 00:04:12.590
The southeast of
the United States

00:04:12.590 --> 00:04:15.460
as well, where the TVA, the
Tennessee Valley Authority,

00:04:15.460 --> 00:04:18.570
has very low-priced
nuclear and coal power.

00:04:18.570 --> 00:04:21.380
So these are examples of where
price might be below cost.

00:04:21.380 --> 00:04:23.880
We're going to get more into
that over the course of today's

00:04:23.880 --> 00:04:26.120
lecture, because
there are some very

00:04:26.120 --> 00:04:30.160
interesting geopolitical
debates occurring right now.

00:04:30.160 --> 00:04:34.540
Oftentimes the two sides are
very staunch in their positions

00:04:34.540 --> 00:04:37.140
and there isn't much nuance,
there isn't much shade of gray,

00:04:37.140 --> 00:04:39.420
there aren't many rational
arguments presented.

00:04:39.420 --> 00:04:42.050
And instead, we're going to
be diving into some of that,

00:04:42.050 --> 00:04:44.870
discussing the nuance
over today's lecture.

00:04:44.870 --> 00:04:47.370
Let's dive into cost first up.

00:04:47.370 --> 00:04:50.520
This is a paper that I
presented already in class.

00:04:50.520 --> 00:04:53.940
I've also steered some of
the project groups toward it.

00:04:53.940 --> 00:04:56.680
This is a proceeding
back in-- whoa.

00:04:56.680 --> 00:04:57.830
This wasn't 2009.

00:04:57.830 --> 00:04:58.440
My apologies.

00:04:58.440 --> 00:05:00.730
This is 2003.

00:05:00.730 --> 00:05:03.300
This was presented at
the 3rd World Conference

00:05:03.300 --> 00:05:06.850
of Photovoltaic Energy
Conversion by Tom Surek,

00:05:06.850 --> 00:05:10.040
presenting a very
simple cost model,

00:05:10.040 --> 00:05:13.310
if you will, for PV, more
specifically the impact

00:05:13.310 --> 00:05:14.455
of efficiency on cost.

00:05:14.455 --> 00:05:16.400
And by no means was
this the first time

00:05:16.400 --> 00:05:18.700
that something like this
had ever been presented,

00:05:18.700 --> 00:05:21.270
but it was a nice
summary of the work

00:05:21.270 --> 00:05:25.030
to date, highlighting
several, I would say,

00:05:25.030 --> 00:05:27.050
key levers, cost levers.

00:05:27.050 --> 00:05:30.170
Efficiency-- that's the
solar conversion efficiency.

00:05:30.170 --> 00:05:33.270
Processing costs-- that's
the materials and processing

00:05:33.270 --> 00:05:37.520
costs for the module in
dollars per meter squared.

00:05:37.520 --> 00:05:40.090
The manufacturing
yield-- that means out

00:05:40.090 --> 00:05:42.430
of, say, 100 cells into
your manufacturing line,

00:05:42.430 --> 00:05:44.180
how many make it through
to the other side

00:05:44.180 --> 00:05:48.110
without breaking
or being discarded

00:05:48.110 --> 00:05:49.920
due to manufacturing defects?

00:05:49.920 --> 00:05:52.832
Capital equipment cost-- that's
depreciated over several years,

00:05:52.832 --> 00:05:54.290
meaning you buy
equipment up front,

00:05:54.290 --> 00:05:56.330
but then due to
financial gimmicks,

00:05:56.330 --> 00:05:59.860
you're allowed to
allow that cost

00:05:59.860 --> 00:06:02.100
to hit your books over an
extended period of time,

00:06:02.100 --> 00:06:04.080
not all at once upfront.

00:06:04.080 --> 00:06:07.070
Overhead, and so forth--
overhead being the health

00:06:07.070 --> 00:06:10.010
insurance, if it is
paid to the workers,

00:06:10.010 --> 00:06:13.860
and, of course, R&D
and the CEO's salary,

00:06:13.860 --> 00:06:16.520
and so forth could be lumped in.

00:06:16.520 --> 00:06:19.590
So this is a very simple
way of estimating cost.

00:06:19.590 --> 00:06:24.660
It's a linear equation,
a direct relationship.

00:06:24.660 --> 00:06:28.680
What is, I would
say, the economics,

00:06:28.680 --> 00:06:31.550
or a more sophisticated
way of looking at cost?

00:06:31.550 --> 00:06:36.090
Other than just saying it's
the dollars per watt-peak,

00:06:36.090 --> 00:06:38.491
you would look at it in terms
of cents per kilowatt hour.

00:06:38.491 --> 00:06:38.990
Right?

00:06:38.990 --> 00:06:40.590
You would look at
it in terms of,

00:06:40.590 --> 00:06:42.715
how much do you pay for
your electricity coming out

00:06:42.715 --> 00:06:43.270
of the wall?

00:06:43.270 --> 00:06:45.370
Or in this case,
out of the panels?

00:06:45.370 --> 00:06:48.260
What would factor
in to what is called

00:06:48.260 --> 00:06:49.845
the levelized cost
of electricity,

00:06:49.845 --> 00:06:52.220
when you're actually calculating
cents per kilowatt hour?

00:06:52.220 --> 00:06:55.150
How would you convert
dollars per watt-peak-- OK, I

00:06:55.150 --> 00:06:58.090
know how many dollars it
took to manufacture this.

00:06:58.090 --> 00:07:02.390
I can also depreciate
my equipment costs

00:07:02.390 --> 00:07:05.310
over several years
to manufacture this.

00:07:05.310 --> 00:07:07.900
How would I go from dollars
per watt-peak into cents

00:07:07.900 --> 00:07:09.550
per kilowatt hour?

00:07:09.550 --> 00:07:11.300
We all agree that cents
per kilowatt hours

00:07:11.300 --> 00:07:12.600
is the metric of
importance, right?

00:07:12.600 --> 00:07:14.433
That's what we pay on
our electricity bills,

00:07:14.433 --> 00:07:15.739
or at least some of us do.

00:07:15.739 --> 00:07:17.280
So when we pay our
electricity bills,

00:07:17.280 --> 00:07:19.672
we're paying in cents per
kilowatt hour from the grid.

00:07:19.672 --> 00:07:21.380
And when we manufacture
our solar panels,

00:07:21.380 --> 00:07:22.713
we pay in dollars per watt-peak.

00:07:22.713 --> 00:07:23.830
Let's start simple.

00:07:23.830 --> 00:07:27.430
Why is dollars per
watt-peak at all useful?

00:07:27.430 --> 00:07:29.610
It's so far removed
from cents per kilowatt

00:07:29.610 --> 00:07:32.200
hour that it almost seems
an artificial metric.

00:07:32.200 --> 00:07:35.310
Why, again, do we use
dollars per watt-peak?

00:07:35.310 --> 00:07:35.810
Jessie?

00:07:35.810 --> 00:07:39.912
AUDIENCE: Because most
coal-based or most

00:07:39.912 --> 00:07:44.892
fossil fuel-based electricity
is based on a capacity factor,

00:07:44.892 --> 00:07:46.844
which is measured in kilowatts.

00:07:46.844 --> 00:07:48.760
PROFESSOR: That's a good
way of looking at it.

00:07:48.760 --> 00:07:53.750
And then the capacity factor of
solar would be based on what?

00:07:53.750 --> 00:07:56.940
On the solar resource
locally, right?

00:07:56.940 --> 00:07:58.881
And that might vary from
location to location.

00:07:58.881 --> 00:07:59.380
OK.

00:07:59.380 --> 00:08:01.330
So what dollars per
watt-peak allows you to do

00:08:01.330 --> 00:08:04.930
is, given a rated nameplate
capacity, you can calculate,

00:08:04.930 --> 00:08:06.930
based on the solar
resource locally,

00:08:06.930 --> 00:08:10.720
how much energy will be produced
over a certain period of time.

00:08:10.720 --> 00:08:12.720
And then from that, you
can calculate your cents

00:08:12.720 --> 00:08:16.290
per kilowatt hour, because now
we're converting from power,

00:08:16.290 --> 00:08:20.170
or rated nameplate power,
into energy, which we can use,

00:08:20.170 --> 00:08:23.080
and which has economic value.

00:08:23.080 --> 00:08:28.220
So there is a rationale, then,
for giving nameplate capacity

00:08:28.220 --> 00:08:29.750
in terms of watt-peak.

00:08:29.750 --> 00:08:32.720
In other words, rating a factory
in terms of megawatts per year

00:08:32.720 --> 00:08:35.340
or gigawatts per year produced.

00:08:35.340 --> 00:08:38.030
That means that each module
that goes on to the cell tester

00:08:38.030 --> 00:08:40.274
is rated, and
there's an estimate

00:08:40.274 --> 00:08:42.190
based on the cumulative
module production what

00:08:42.190 --> 00:08:45.120
the total watt-peak output
of that factory was.

00:08:45.120 --> 00:08:47.540
And then depending on where
those modules go in the world,

00:08:47.540 --> 00:08:49.721
they might produce
different amounts of energy.

00:08:49.721 --> 00:08:51.637
If you take those same
models and install them

00:08:51.637 --> 00:08:53.053
in Alaska or
Arizona, you're going

00:08:53.053 --> 00:08:55.130
to get widely varying
energy outputs.

00:08:55.130 --> 00:08:55.700
OK.

00:08:55.700 --> 00:08:57.430
So then, how do we transition?

00:08:57.430 --> 00:09:00.310
We have the dollars
per watt-peak.

00:09:00.310 --> 00:09:03.340
We have to know the local
solar insulation that

00:09:03.340 --> 00:09:05.980
would allow us to calculate
what the cents per kilowatt hour

00:09:05.980 --> 00:09:08.940
would be, assuming a
certain cost of capital.

00:09:08.940 --> 00:09:11.820
We have to buy those
panels up front.

00:09:11.820 --> 00:09:15.510
You have to buy from me a huge
number of panels, which are

00:09:15.510 --> 00:09:16.970
going to last for 20, 25 years.

00:09:16.970 --> 00:09:18.180
I'll guarantee it.

00:09:18.180 --> 00:09:20.440
But you have to front
that money up front,

00:09:20.440 --> 00:09:23.790
which means that you need to
lend that money from a bank

00:09:23.790 --> 00:09:26.250
or from a financial
institution, and then you'll

00:09:26.250 --> 00:09:28.399
be paying a certain amount
of interest every year.

00:09:28.399 --> 00:09:30.940
And it's the spread, it's the
difference between the interest

00:09:30.940 --> 00:09:34.060
payments and the money
saved that's going

00:09:34.060 --> 00:09:35.920
to turn your profit.

00:09:35.920 --> 00:09:38.420
And that's what's called
the rate of return

00:09:38.420 --> 00:09:39.750
of your investment.

00:09:39.750 --> 00:09:42.210
And there's also a payback
period, or a monetary payback

00:09:42.210 --> 00:09:45.980
period, over which you're
not going to be making money

00:09:45.980 --> 00:09:48.630
on average, but after the
payback period is finished,

00:09:48.630 --> 00:09:52.060
then your solar array
will be a money press.

00:09:52.060 --> 00:09:53.130
It'll be printing money.

00:09:53.130 --> 00:09:55.460
And so over the entire
lifetime of the investment,

00:09:55.460 --> 00:09:58.320
you can calculate an
average rate of return.

00:09:58.320 --> 00:09:59.020
OK.

00:09:59.020 --> 00:10:04.200
So given that setup,
we're going to be talking

00:10:04.200 --> 00:10:09.610
about some incentive
mechanisms and, today, really,

00:10:09.610 --> 00:10:11.540
really simple cost model.

00:10:11.540 --> 00:10:15.250
More on Thursday
when we get back

00:10:15.250 --> 00:10:19.470
to this, how we calculate cost
in a more sophisticated manner.

00:10:19.470 --> 00:10:23.920
I want to introduce this
concept very simply at first,

00:10:23.920 --> 00:10:26.490
just because there are
some parameters here

00:10:26.490 --> 00:10:29.260
in the sensitivity analysis
that overwhelm all the others.

00:10:29.260 --> 00:10:32.140
And they can be very simply
seen an equation like that right

00:10:32.140 --> 00:10:33.070
up there.

00:10:33.070 --> 00:10:34.700
And one of them is efficiency.

00:10:34.700 --> 00:10:37.990
The processing costs
also matter quite a bit.

00:10:37.990 --> 00:10:38.850
OK.

00:10:38.850 --> 00:10:42.150
Inside of processing
costs, you also

00:10:42.150 --> 00:10:45.950
have labor and
commodity materials.

00:10:45.950 --> 00:10:48.260
And those can vary significantly
from region to region.

00:10:51.540 --> 00:10:56.310
There are a number of
assessments out there.

00:10:56.310 --> 00:11:00.250
I'm going to be leading us into
the current big debate which

00:11:00.250 --> 00:11:03.710
has exploded in Washington
DC, the case of Solyndra,

00:11:03.710 --> 00:11:11.710
and then SolarWorld filing the
complaint with the US Commerce

00:11:11.710 --> 00:11:14.260
Department.

00:11:14.260 --> 00:11:16.110
I'll be easing us into
this whole question

00:11:16.110 --> 00:11:21.130
of US and foreign manufacturing,
using that more as a hook

00:11:21.130 --> 00:11:24.070
to get us interested in the
overall topics of the day.

00:11:24.070 --> 00:11:25.860
And then toward the
end, hopefully you'll

00:11:25.860 --> 00:11:28.330
be able to see, with
more shades of gray,

00:11:28.330 --> 00:11:30.350
what exactly is going
on and perhaps formulate

00:11:30.350 --> 00:11:31.910
opinions of your own.

00:11:31.910 --> 00:11:34.700
So where is PV cost actually at?

00:11:34.700 --> 00:11:39.580
Where do the current cost
numbers currently stand?

00:11:39.580 --> 00:11:42.140
It is extremely difficult
to get at true cost.

00:11:42.140 --> 00:11:42.710
Price?

00:11:42.710 --> 00:11:43.572
Price is easy.

00:11:43.572 --> 00:11:45.030
You go out there,
probe the market,

00:11:45.030 --> 00:11:47.079
see what people are
willing to sell panels at,

00:11:47.079 --> 00:11:49.370
maybe send a few emails to
a few companies saying, hey,

00:11:49.370 --> 00:11:51.417
I want to install
100 of your panels.

00:11:51.417 --> 00:11:53.250
How much are you willing
to sell them to me?

00:11:53.250 --> 00:11:54.050
This is what people do.

00:11:54.050 --> 00:11:54.840
They probe the market.

00:11:54.840 --> 00:11:55.381
They test it.

00:11:55.381 --> 00:11:57.640
So price is fairly
easy to gauge.

00:11:57.640 --> 00:11:59.420
But manufacturing cost?

00:11:59.420 --> 00:12:01.270
If I go up to you
and say, hey, can you

00:12:01.270 --> 00:12:03.440
tell me how much it costs
to manufacture panels

00:12:03.440 --> 00:12:04.419
in your factory?

00:12:04.419 --> 00:12:06.460
Would you be willing to
give me that information?

00:12:06.460 --> 00:12:08.240
They'd probably say no.

00:12:08.240 --> 00:12:08.990
Probably say no.

00:12:08.990 --> 00:12:11.060
Definitely to me,
as a professor.

00:12:11.060 --> 00:12:12.964
Maybe you as students,
they might show you

00:12:12.964 --> 00:12:15.380
a little bit more of how it
actually works under the hood,

00:12:15.380 --> 00:12:17.900
since many people believe
in the educational mission.

00:12:17.900 --> 00:12:20.750
But, by and large, people are
fairly secretive about this,

00:12:20.750 --> 00:12:23.000
because they realize that
their stock price is heavily

00:12:23.000 --> 00:12:24.220
dependent on market perception.

00:12:24.220 --> 00:12:26.345
And if I go out there and
write an article and say,

00:12:26.345 --> 00:12:28.770
hey, your company
produces panels

00:12:28.770 --> 00:12:30.312
at twice the cost
of yours, investors

00:12:30.312 --> 00:12:32.728
are going to flee your company,
especially if it comes out

00:12:32.728 --> 00:12:34.350
of a university like MIT.

00:12:34.350 --> 00:12:37.620
And so there are big
repercussions associated

00:12:37.620 --> 00:12:40.430
with the divulgation
of cost numbers,

00:12:40.430 --> 00:12:42.900
and that's why it's very hard
to get to the bottom of it.

00:12:42.900 --> 00:12:45.290
What you find most
often are aggregators.

00:12:45.290 --> 00:12:49.030
These are consultancies that
work with several companies.

00:12:49.030 --> 00:12:51.920
They may be, for example,
photon consulting

00:12:51.920 --> 00:12:53.830
across the river in Boston.

00:12:53.830 --> 00:12:55.960
It might be Greentech
Media consulting

00:12:55.960 --> 00:12:58.867
branch, which is located here
in Cambridge, Massachusetts.

00:12:58.867 --> 00:13:01.200
There are many consulting
companies, and several of them

00:13:01.200 --> 00:13:03.930
based here in the Boston area.

00:13:03.930 --> 00:13:06.660
And these consulting
companies work with several PV

00:13:06.660 --> 00:13:08.720
manufacturers and,
over time, begin

00:13:08.720 --> 00:13:11.520
aggregating data and
presenting market trends,

00:13:11.520 --> 00:13:13.015
generalized market trends.

00:13:13.015 --> 00:13:15.620
A few of them single
out specific companies,

00:13:15.620 --> 00:13:17.700
but most of them just
aggregate the data

00:13:17.700 --> 00:13:19.690
and present general trends.

00:13:19.690 --> 00:13:24.510
I presented to you this,
a few articles as of late.

00:13:24.510 --> 00:13:28.860
These are relatively recent--
within the last three to four

00:13:28.860 --> 00:13:31.260
years-- and because
there has been

00:13:31.260 --> 00:13:33.990
so much change over the
last three to four years

00:13:33.990 --> 00:13:36.620
within the PV market, even
an article three years ago

00:13:36.620 --> 00:13:38.540
is highly outdated.

00:13:38.540 --> 00:13:41.530
So you'll want information
that's within the last year

00:13:41.530 --> 00:13:44.020
if you're going to be
using cost and price

00:13:44.020 --> 00:13:45.560
information for your projects.

00:13:45.560 --> 00:13:47.939
Even within the last few months.

00:13:47.939 --> 00:13:49.980
And if any of you have
any doubts in that regard,

00:13:49.980 --> 00:13:51.530
you can come talk to me.

00:13:51.530 --> 00:13:54.620
So the fully loaded
module manufacturing cost

00:13:54.620 --> 00:13:56.340
is shown here.

00:13:56.340 --> 00:13:57.920
An estimate, again.

00:13:57.920 --> 00:14:02.300
An aggregation based on access
to multiple companies' data.

00:14:02.300 --> 00:14:05.000
Polysilicon-- this, of course,
being crystalline silicon

00:14:05.000 --> 00:14:07.750
technology, which
accounts for about 85%

00:14:07.750 --> 00:14:08.970
of the current market.

00:14:08.970 --> 00:14:12.622
Polysilicon in blue,
depreciation in red.

00:14:12.622 --> 00:14:13.455
What's depreciation?

00:14:16.580 --> 00:14:21.327
AUDIENCE: Is that the
discounting of the capital

00:14:21.327 --> 00:14:22.410
that we just talked about?

00:14:22.410 --> 00:14:23.750
PROFESSOR: Exactly.

00:14:23.750 --> 00:14:24.280
Exactly.

00:14:24.280 --> 00:14:27.510
So what that means is,
I borrow a lot of money

00:14:27.510 --> 00:14:30.560
to manufacture these panels,
to buy the equipment,

00:14:30.560 --> 00:14:33.230
to buy the factory
and set everything up.

00:14:33.230 --> 00:14:36.570
And then I have to pay
interest on my loans.

00:14:36.570 --> 00:14:40.260
I have to pay interest on
the loans, and what I do is,

00:14:40.260 --> 00:14:42.360
I start writing
off the equipment

00:14:42.360 --> 00:14:44.740
as a loss to the company.

00:14:44.740 --> 00:14:48.480
So that equipment is going to
be useful over a certain period

00:14:48.480 --> 00:14:51.050
of time, and then we can
assume that it's outdated

00:14:51.050 --> 00:14:52.925
and that I'll need to
buy the next generation

00:14:52.925 --> 00:14:54.150
of manufacturing equipment.

00:14:54.150 --> 00:14:56.930
So I can begin writing off
the value of that equipment

00:14:56.930 --> 00:14:59.360
over a certain schedule, over
a certain period of years.

00:14:59.360 --> 00:15:01.420
Five, seven, depending
on the company,

00:15:01.420 --> 00:15:04.880
depending on the way they're--
I don't want to say cooking

00:15:04.880 --> 00:15:08.139
their books, but the way
they're manipulating the numbers

00:15:08.139 --> 00:15:09.180
in the accounting sector.

00:15:09.180 --> 00:15:12.490
So the depreciation varies.

00:15:12.490 --> 00:15:16.060
And it varies because the
interest rate at which you

00:15:16.060 --> 00:15:19.500
acquire the loan,
and the interest

00:15:19.500 --> 00:15:24.090
rate-- or the inflation rate
of the particular currency

00:15:24.090 --> 00:15:26.380
in question, is different
in different regions.

00:15:26.380 --> 00:15:28.356
So when you're
calculating depreciation

00:15:28.356 --> 00:15:29.980
both of those things
matter, and that's

00:15:29.980 --> 00:15:32.560
why those numbers can
vary from place to place.

00:15:32.560 --> 00:15:33.420
Materials.

00:15:33.420 --> 00:15:34.080
Materials cost.

00:15:34.080 --> 00:15:35.690
These are the materials
used to manufacture

00:15:35.690 --> 00:15:37.731
the module, typically
called commodity materials.

00:15:37.731 --> 00:15:41.050
You would look at them as the
extruded aluminum components,

00:15:41.050 --> 00:15:44.919
the glass in the front side,
the encapsulant materials,

00:15:44.919 --> 00:15:45.460
and so forth.

00:15:45.460 --> 00:15:48.050
All the fancy stuff that we
saw when we visited Fraunhofer

00:15:48.050 --> 00:15:50.877
last week.

00:15:50.877 --> 00:15:51.710
Let's see what else.

00:15:51.710 --> 00:15:52.790
We have labor.

00:15:52.790 --> 00:15:55.020
That's pretty
straightforward to see.

00:15:55.020 --> 00:15:58.450
If we're manufacturing in China,
we are looking at a labor rate,

00:15:58.450 --> 00:16:02.890
could be as low-- base
labor without adding housing

00:16:02.890 --> 00:16:05.420
and so forth-- the base labor
rate could be as low as $2.75

00:16:05.420 --> 00:16:07.710
an hour in US dollars.

00:16:07.710 --> 00:16:10.340
And in US dollars in
the United States,

00:16:10.340 --> 00:16:13.460
we could be looking at
labor rates of somewhere

00:16:13.460 --> 00:16:16.380
in the range of $16
an hour before you

00:16:16.380 --> 00:16:19.830
start adding in Social Security
and benefits and so forth.

00:16:19.830 --> 00:16:23.310
And there may be
different levels

00:16:23.310 --> 00:16:25.480
of automation in the two
different places, which

00:16:25.480 --> 00:16:29.260
shifts costs from labor
into capital equipment.

00:16:29.260 --> 00:16:32.150
So if you realize that you
have a much higher labor cost,

00:16:32.150 --> 00:16:35.315
you might want to buy more
robots to do the manufacturing.

00:16:35.315 --> 00:16:37.440
And vice versa, if you're
in China and you realize,

00:16:37.440 --> 00:16:39.290
oh my goodness,
our labor rates are

00:16:39.290 --> 00:16:41.740
increasing almost
exponentially, definitely

00:16:41.740 --> 00:16:43.730
super linearly with time.

00:16:43.730 --> 00:16:46.045
As the country
takes off and there

00:16:46.045 --> 00:16:50.440
is inflation, wage
inflation, a company that

00:16:50.440 --> 00:16:53.439
is trying to project forward
five or seven years might say,

00:16:53.439 --> 00:16:55.230
well goodness, it
doesn't make sense for me

00:16:55.230 --> 00:16:56.980
to flood my manufacturing
line with people

00:16:56.980 --> 00:16:59.290
right now, because that
manufacturing line still has

00:16:59.290 --> 00:17:01.530
to make a profit in five years.

00:17:01.530 --> 00:17:03.610
And so I'm going to
buy more robots now.

00:17:03.610 --> 00:17:05.380
Even though I might
not need it today,

00:17:05.380 --> 00:17:06.794
I might need in five years.

00:17:06.794 --> 00:17:08.460
And so there's a bit
of risk calculation

00:17:08.460 --> 00:17:11.250
that gets thrown
into this as well.

00:17:11.250 --> 00:17:12.359
Utilities and overhead.

00:17:12.359 --> 00:17:13.609
That's pretty straightforward.

00:17:13.609 --> 00:17:16.770
For looking at utilities, that
means the electricity, mostly,

00:17:16.770 --> 00:17:18.349
to run the lines.

00:17:18.349 --> 00:17:22.020
It could also be water which
is used in the manufacturing

00:17:22.020 --> 00:17:22.520
process.

00:17:25.450 --> 00:17:27.849
Note the caveats in
several of these studies,

00:17:27.849 --> 00:17:30.160
especially the one out of
Lawrence Berkeley National

00:17:30.160 --> 00:17:31.790
Laboratory.

00:17:31.790 --> 00:17:36.460
Manufacturing cost, you can do
a number of things with that.

00:17:36.460 --> 00:17:39.260
You can assume that you're
buying your polysilicon,

00:17:39.260 --> 00:17:42.160
you're buying the polysilicon,
manufacturing cells, modules,

00:17:42.160 --> 00:17:43.030
and systems.

00:17:43.030 --> 00:17:45.820
Or you can assume that
you're buying the cells,

00:17:45.820 --> 00:17:48.271
and the cells are commodity
products equal in price

00:17:48.271 --> 00:17:50.020
throughout the world,
and that you're just

00:17:50.020 --> 00:17:51.600
manufacturing the modules.

00:17:51.600 --> 00:17:54.500
So you can cut costs
in many different ways,

00:17:54.500 --> 00:17:56.380
depending on what
assumptions you make.

00:17:56.380 --> 00:17:58.972
The price of the cell on
the international market

00:17:58.972 --> 00:18:00.430
may be very different
than the cost

00:18:00.430 --> 00:18:03.750
to manufacture that cell
in your particular country.

00:18:03.750 --> 00:18:05.570
And cost is almost
one of those things

00:18:05.570 --> 00:18:07.410
that you can torture
until it tells you

00:18:07.410 --> 00:18:09.620
the story you want it to tell.

00:18:09.620 --> 00:18:12.685
It's one of those
things where, if you

00:18:12.685 --> 00:18:14.890
did these numbers a little
bit differently and said,

00:18:14.890 --> 00:18:19.990
OK, we're buying cells and just
making modules, and then adding

00:18:19.990 --> 00:18:22.180
shipment fees-- let's say
transport fees from China

00:18:22.180 --> 00:18:26.450
to the United States-- China
and US might look almost equal.

00:18:26.450 --> 00:18:28.714
Whereas if you integrate
over the entire supply chain

00:18:28.714 --> 00:18:31.130
and say, OK, I'm going to be
manufacturing my polysilicon,

00:18:31.130 --> 00:18:35.170
then my wafers, then my cells,
and finally the modules,

00:18:35.170 --> 00:18:36.920
then you might start
seeing some disparity

00:18:36.920 --> 00:18:39.425
based on these other
parameters shown here.

00:18:39.425 --> 00:18:44.520
The point being, be careful
when you see a cost assessment.

00:18:44.520 --> 00:18:46.990
Probe their base
assumptions and try

00:18:46.990 --> 00:18:49.880
to understand what their
biases and motivations were

00:18:49.880 --> 00:18:53.180
for presenting that
particular comparison.

00:18:53.180 --> 00:18:56.490
Especially nowadays, where
you have more and more parties

00:18:56.490 --> 00:18:58.290
with vested interests
in presenting

00:18:58.290 --> 00:19:00.530
one story or another.

00:19:00.530 --> 00:19:02.330
We'll get to that
in a few slides.

00:19:02.330 --> 00:19:03.920
Experience learning curve.

00:19:03.920 --> 00:19:05.760
Now we're getting into price.

00:19:05.760 --> 00:19:08.870
We're venturing beyond
cost and into the regime

00:19:08.870 --> 00:19:12.020
of price, which is definitely
more easily measurable.

00:19:12.020 --> 00:19:16.410
And that's why we have some
fairly good data going back

00:19:16.410 --> 00:19:17.880
several decades.

00:19:17.880 --> 00:19:21.800
And you can see here,
this is cumulative sales

00:19:21.800 --> 00:19:22.790
in gigawatt peak.

00:19:22.790 --> 00:19:24.850
That means the cumulative
number of widgets,

00:19:24.850 --> 00:19:28.290
in, this case, watt-peaks,
produced by solar industries

00:19:28.290 --> 00:19:29.370
worldwide.

00:19:29.370 --> 00:19:33.550
And this is a global average
module selling price.

00:19:33.550 --> 00:19:35.640
Not the system, the module.

00:19:35.640 --> 00:19:38.090
So not the balance of
system installation, labor,

00:19:38.090 --> 00:19:41.360
and so forth, but just the
manufacturing of the module.

00:19:41.360 --> 00:19:43.900
And what we can see
here is a general trend

00:19:43.900 --> 00:19:46.060
over time with a
decreasing price.

00:19:46.060 --> 00:19:46.760
That's good.

00:19:46.760 --> 00:19:48.430
Sorry, decreasing
price with time.

00:19:48.430 --> 00:19:50.350
That's excellent.

00:19:50.350 --> 00:19:53.150
What is that called?

00:19:53.150 --> 00:19:55.730
This curve here, plotted
in a log-log scale,

00:19:55.730 --> 00:19:59.890
where you have a line through
the cumulative manufacturing

00:19:59.890 --> 00:20:02.330
production versus price.

00:20:02.330 --> 00:20:03.466
What is that curve called?

00:20:03.466 --> 00:20:05.340
AUDIENCE: Is it an
experience learning curve?

00:20:05.340 --> 00:20:07.214
PROFESSOR: It's an
experience learning curve.

00:20:07.214 --> 00:20:10.510
And you'll get something similar
for any high-tech product--

00:20:10.510 --> 00:20:13.550
computers, toasters-- as long
as the product isn't changing

00:20:13.550 --> 00:20:14.710
significantly with time.

00:20:14.710 --> 00:20:16.931
And even sometimes
if they are, as is

00:20:16.931 --> 00:20:18.180
the case with the solar panel.

00:20:18.180 --> 00:20:20.720
Solar panels back in the
1960s, or the 1970s, rather,

00:20:20.720 --> 00:20:22.511
looked very different
than the solar panels

00:20:22.511 --> 00:20:24.830
today in terms of the materials
and the processes used.

00:20:24.830 --> 00:20:26.880
And it all falls along this
very interesting experience

00:20:26.880 --> 00:20:27.600
learning curve.

00:20:27.600 --> 00:20:30.200
So it's extremely
tempting to say, oh well,

00:20:30.200 --> 00:20:33.550
what price do we need
to reach to be cost

00:20:33.550 --> 00:20:35.360
competitive-- or
competitive, let's say,

00:20:35.360 --> 00:20:36.571
with bulk electricity?

00:20:36.571 --> 00:20:38.820
We need to reach a price of,
say, $0.50 per watt-peak?

00:20:38.820 --> 00:20:40.050
Oh, easy.

00:20:40.050 --> 00:20:41.550
We'll just project
forward and we'll

00:20:41.550 --> 00:20:44.160
see how much cumulative
production is needed,

00:20:44.160 --> 00:20:45.830
and then we'll
subsidize until we get

00:20:45.830 --> 00:20:47.080
to that cumulative production.

00:20:47.080 --> 00:20:49.530
And bingo, voila,
it'll happen by magic.

00:20:49.530 --> 00:20:52.930
Well, the reality is that
each little bump here

00:20:52.930 --> 00:20:55.407
along the learning curve was
some-- if you look closely,

00:20:55.407 --> 00:20:57.240
you can kind of see
these little bumps here.

00:20:57.240 --> 00:20:59.490
There were many traumatic
events within the industry

00:20:59.490 --> 00:21:04.670
that forced people to innovate,
to produce better technology,

00:21:04.670 --> 00:21:07.580
whether it's the technology
itself, something designed here

00:21:07.580 --> 00:21:09.820
in the laboratory
at MIT or Harvard,

00:21:09.820 --> 00:21:12.800
or whether it's something
innovated on the manufacturing

00:21:12.800 --> 00:21:14.550
line where they
realized, oh, this

00:21:14.550 --> 00:21:16.966
is a more efficient way of
manufacturing the solar panels.

00:21:16.966 --> 00:21:19.120
It's cheaper.

00:21:19.120 --> 00:21:21.615
This little bump
right here, boom.

00:21:21.615 --> 00:21:23.865
This represents a period in
which prices actually went

00:21:23.865 --> 00:21:25.870
back up year in and year out.

00:21:25.870 --> 00:21:29.130
What could cause
prices to go up?

00:21:29.130 --> 00:21:31.686
What are some of the
motivations for that?

00:21:31.686 --> 00:21:32.810
I mean, scale's increasing.

00:21:32.810 --> 00:21:33.990
The market's growing.

00:21:33.990 --> 00:21:35.250
It's not shrinking at all.

00:21:35.250 --> 00:21:38.382
It's not like the points went
back here as they went up.

00:21:38.382 --> 00:21:40.840
So the market continued to
grow, the manufacturing capacity

00:21:40.840 --> 00:21:42.756
continued to grow, but
the price went back up.

00:21:42.756 --> 00:21:44.280
What could have caused that?

00:21:44.280 --> 00:21:46.630
AUDIENCE: Demand
outpaces supply?

00:21:46.630 --> 00:21:48.520
PROFESSOR: Demand
outpaces supply.

00:21:48.520 --> 00:21:49.030
Exactly.

00:21:49.030 --> 00:21:52.020
So in this specific
case, what happened

00:21:52.020 --> 00:21:54.890
was the polysilicon
feedstock, which

00:21:54.890 --> 00:21:57.370
is the input material
into this entire process,

00:21:57.370 --> 00:21:59.010
was in short supply.

00:21:59.010 --> 00:22:02.640
It takes about, in those days
it took about 24 to 36 months

00:22:02.640 --> 00:22:04.180
to get a new plant online.

00:22:04.180 --> 00:22:06.960
Long lead time and billions
of dollars of investment.

00:22:06.960 --> 00:22:09.640
And so the polysilicon
suppliers didn't really

00:22:09.640 --> 00:22:13.230
want to invest unless they knew
photovoltaics was for real.

00:22:13.230 --> 00:22:15.280
And the PV industry
had exhausted

00:22:15.280 --> 00:22:18.030
the elasticity of
the supply market

00:22:18.030 --> 00:22:20.430
in the polysilicon business.

00:22:20.430 --> 00:22:22.770
And polysilicon suppliers
looked at the situation

00:22:22.770 --> 00:22:25.610
and said, well, let's let
prices go up a little bit.

00:22:25.610 --> 00:22:28.180
It can't hurt us too bad.

00:22:28.180 --> 00:22:30.740
We've been starved for
several years because

00:22:30.740 --> 00:22:32.910
of low polysilicon
feedstock prices.

00:22:32.910 --> 00:22:35.480
Let's let this increased
demand kind of push prices

00:22:35.480 --> 00:22:37.895
up a little bit before we
really decide what to do.

00:22:37.895 --> 00:22:39.270
And by the time
they decided what

00:22:39.270 --> 00:22:42.230
to do, there was a lot of
them getting in the market

00:22:42.230 --> 00:22:44.200
all at once, which
had this effect.

00:22:44.200 --> 00:22:44.770
Boom.

00:22:44.770 --> 00:22:47.500
Now, it's not only the
polysilicon, but also

00:22:47.500 --> 00:22:50.270
the cell manufacturers,
the wafer manufacturers,

00:22:50.270 --> 00:22:54.900
that expanded their capacity
during this time between 2007

00:22:54.900 --> 00:22:55.840
and 2010.

00:22:55.840 --> 00:22:59.420
Now today's most
recent price point,

00:22:59.420 --> 00:23:05.060
as I saw it on one of our
more trusted websites,

00:23:05.060 --> 00:23:09.880
put us down at around
$1.03, $1.05 per watt-peak.

00:23:09.880 --> 00:23:10.730
So we're down here.

00:23:10.730 --> 00:23:13.310
We're well below the
historical average trend.

00:23:13.310 --> 00:23:17.090
We're in an oversupply
condition right now.

00:23:17.090 --> 00:23:20.040
Why did the oversupply
condition come about?

00:23:20.040 --> 00:23:23.320
Well, partly because of
this undersupply condition,

00:23:23.320 --> 00:23:25.060
many people saw an
opportunity and said,

00:23:25.060 --> 00:23:26.970
well, we can
address that demand.

00:23:26.970 --> 00:23:29.280
We can grow in this
industry right here.

00:23:29.280 --> 00:23:31.950
To grow in the industry,
you need capital.

00:23:31.950 --> 00:23:35.400
You need access to
finances to expand.

00:23:35.400 --> 00:23:40.030
What happened in 2008 in
Western countries-- and 2008

00:23:40.030 --> 00:23:45.410
was more or less when it really
hit the fan, if you will,

00:23:45.410 --> 00:23:48.440
in the United States
in particular.

00:23:48.440 --> 00:23:50.510
What happened in
the capital markets?

00:23:53.060 --> 00:23:56.896
AUDIENCE: Capital was severely
restricted becase we're not

00:23:56.896 --> 00:23:58.390
providing a lot bones.

00:23:58.390 --> 00:23:59.220
PROFESSOR: Exactly.

00:23:59.220 --> 00:24:02.200
We had the financial
crisis here in the US.

00:24:02.200 --> 00:24:08.100
So in the United States, there
was a pull back of lending.

00:24:08.100 --> 00:24:10.485
The government stepped
in shortly thereafter,

00:24:10.485 --> 00:24:12.360
realizing that this was
going to be an issue.

00:24:12.360 --> 00:24:13.720
There was the ARRA.

00:24:13.720 --> 00:24:15.670
Does anybody know what that is?

00:24:15.670 --> 00:24:17.860
It's not the American
Association of Retired People.

00:24:17.860 --> 00:24:19.630
This is very different.

00:24:19.630 --> 00:24:22.330
ARRA is the American Recovery
and Reinvestment Act, right?

00:24:22.330 --> 00:24:23.950
This was the Stimulus Act.

00:24:23.950 --> 00:24:28.935
And this act actually did
inject a lot of money,

00:24:28.935 --> 00:24:31.970
or a lot of capital, into
renewable energy projects that

00:24:31.970 --> 00:24:35.850
began hitting in
sometime between 2009,

00:24:35.850 --> 00:24:40.790
let's say, at the
beginning, until 2011.

00:24:40.790 --> 00:24:43.740
In this period right over here.

00:24:43.740 --> 00:24:45.182
Anybody hear of
the word Solyndra?

00:24:45.182 --> 00:24:45.682
Yeah?

00:24:45.682 --> 00:24:47.265
You've heard about it?

00:24:47.265 --> 00:24:48.640
Solyndra was one
of the companies

00:24:48.640 --> 00:24:52.672
that received funding
under the ARRA, or "arra."

00:24:52.672 --> 00:24:54.820
Under the Stimulus Act.

00:24:54.820 --> 00:24:57.260
So a lot of
interesting things were

00:24:57.260 --> 00:25:00.890
happening during this brief
little period right here.

00:25:00.890 --> 00:25:06.730
During the mid 2000s, China
was getting a lot of bad rap

00:25:06.730 --> 00:25:09.240
by environmental groups,
and the United States

00:25:09.240 --> 00:25:11.850
as well, for its growing
greenhouse gas emissions.

00:25:11.850 --> 00:25:15.580
There was a growing concern
over greenhouse gases

00:25:15.580 --> 00:25:18.150
culminating in the
Copenhagen Discussions,

00:25:18.150 --> 00:25:20.860
that countries, especially
developing countries,

00:25:20.860 --> 00:25:24.660
had to do more to reduce their
greenhouse gas production.

00:25:24.660 --> 00:25:27.960
And this, of course,
invited a wonderful tug

00:25:27.960 --> 00:25:30.050
of war between a
developing country

00:25:30.050 --> 00:25:32.450
bloc-- the most progressive
of the developing countries,

00:25:32.450 --> 00:25:37.400
let's say, normally described
as BRICO, so Brazil, Russia,

00:25:37.400 --> 00:25:41.640
India, China, and other,
including South Africa--

00:25:41.640 --> 00:25:43.030
and the developed countries.

00:25:43.030 --> 00:25:47.080
And you can see perspectives
from both, and it's wonderful.

00:25:47.080 --> 00:25:50.000
The sophistication
of that debate

00:25:50.000 --> 00:25:54.547
in terms of poli-sci arguments
was just beautiful to watch.

00:25:54.547 --> 00:25:56.380
On one hand you had the
developing countries

00:25:56.380 --> 00:25:58.320
that said, well, wait a second.

00:25:58.320 --> 00:26:00.080
You folks in the
developed world,

00:26:00.080 --> 00:26:02.190
you used a lot of fossil fuels.

00:26:02.190 --> 00:26:04.900
And the majority of
the CO2 emissions

00:26:04.900 --> 00:26:08.380
that have occurred to date
were from developed countries,

00:26:08.380 --> 00:26:09.460
today.

00:26:09.460 --> 00:26:11.680
So it's a bit unfair
that you're asking

00:26:11.680 --> 00:26:14.590
us to reduce our CO2
intensity, because that might

00:26:14.590 --> 00:26:16.280
hamper our own development.

00:26:16.280 --> 00:26:20.220
Why don't you pay
reparations for the CO2

00:26:20.220 --> 00:26:21.830
that you've already
emitted and help

00:26:21.830 --> 00:26:24.145
us decrease our CO2 intensity?

00:26:24.145 --> 00:26:26.520
So there were these beautiful
arguments being constructed

00:26:26.520 --> 00:26:29.010
on both sides of the debate.

00:26:29.010 --> 00:26:31.590
What China decided to do-- so
on the international spectrum,

00:26:31.590 --> 00:26:33.050
not much happened.

00:26:33.050 --> 00:26:36.900
And that's fairly-- sadly,
it's fairly typical of most

00:26:36.900 --> 00:26:40.960
of today's-- I'd say the
larger the body happens to be,

00:26:40.960 --> 00:26:44.860
whether it's the federal
government or the world

00:26:44.860 --> 00:26:47.700
institutions, it seems that
things happen at a much slower

00:26:47.700 --> 00:26:50.220
pace the larger the entity is.

00:26:50.220 --> 00:26:53.510
But at a smaller entity, for
example, the state level, which

00:26:53.510 --> 00:26:55.680
we'll see in a few slides,
in the United States,

00:26:55.680 --> 00:26:57.440
a lot's happening right now.

00:26:57.440 --> 00:26:59.250
And within China's
central government,

00:26:59.250 --> 00:27:01.560
a lot happened in response
to some of that criticism.

00:27:01.560 --> 00:27:04.730
They said well, there's a point.

00:27:04.730 --> 00:27:07.150
More a point that
fossil fuel emissions

00:27:07.150 --> 00:27:09.115
really decreases
our quality of life.

00:27:09.115 --> 00:27:11.240
If you look at some of the
pollution in our cities,

00:27:11.240 --> 00:27:14.320
that's not very becoming.

00:27:14.320 --> 00:27:15.930
We can do better.

00:27:15.930 --> 00:27:17.680
And furthermore,
we can create jobs

00:27:17.680 --> 00:27:19.521
for the people who are
flooding our cities

00:27:19.521 --> 00:27:21.770
from the countryside looking
for economic opportunity.

00:27:21.770 --> 00:27:24.080
We can create new
jobs in this industry.

00:27:24.080 --> 00:27:26.890
And we'll do it
by making capital

00:27:26.890 --> 00:27:30.090
available to this new
nascent industry at a time

00:27:30.090 --> 00:27:32.340
when it's very difficult to
achieve capital or acquire

00:27:32.340 --> 00:27:35.300
capital in the West, in
Europe and the United States.

00:27:35.300 --> 00:27:39.980
And so that's again happening
really in the mid 2000s.

00:27:39.980 --> 00:27:42.660
And what we'll see
in a few slides

00:27:42.660 --> 00:27:45.820
is the massive growth of
the Chinese manufacturing

00:27:45.820 --> 00:27:49.500
market in response to the
availability of capital

00:27:49.500 --> 00:27:50.790
in those countries.

00:27:50.790 --> 00:27:53.110
Let me go back to a
slide that I presented,

00:27:53.110 --> 00:27:54.980
I think it was lecture
number 1, where

00:27:54.980 --> 00:27:57.410
we looked at the
cumulative production of PV

00:27:57.410 --> 00:27:58.880
as a function of year.

00:27:58.880 --> 00:28:02.390
And if we plot this on
the log-linear plot,

00:28:02.390 --> 00:28:07.150
we can read the growth rate
off of the slope of that curve.

00:28:07.150 --> 00:28:12.400
So this was somewhere around
10%, 40%, maybe upwards to 60%,

00:28:12.400 --> 00:28:17.230
depending on what data points
you include in those lines.

00:28:17.230 --> 00:28:19.340
Interestingly,
back in 1990, if we

00:28:19.340 --> 00:28:22.840
look at the distribution
of different technologies,

00:28:22.840 --> 00:28:25.880
multicrystalline silicon was
1/3, single-crystalline silicon

00:28:25.880 --> 00:28:28.660
about 1/3, and thin-film
technology, namely,

00:28:28.660 --> 00:28:30.807
pushed by amorphous
silicon, was about 1/3

00:28:30.807 --> 00:28:32.140
of all manufacturing production.

00:28:34.670 --> 00:28:38.270
As the market evolved--
sorry about this.

00:28:38.270 --> 00:28:38.820
There we go.

00:28:38.820 --> 00:28:40.384
So again, the
different technologies,

00:28:40.384 --> 00:28:42.300
just to situate ourselves,
single-crystalline,

00:28:42.300 --> 00:28:44.020
multicrystalline,
and thin films.

00:28:44.020 --> 00:28:47.050
As the market evolved
going into the 2000s,

00:28:47.050 --> 00:28:51.020
we saw this type
of breakdown occur.

00:28:51.020 --> 00:28:54.470
We have silicon, or
crystalline silicon comprising

00:28:54.470 --> 00:28:59.370
around 85% to 90% of the market,
and thin films not growing

00:28:59.370 --> 00:29:01.790
as fast as crystalline silicon.

00:29:01.790 --> 00:29:02.940
It was still growing.

00:29:02.940 --> 00:29:04.930
The market overall
is going gangbusters,

00:29:04.930 --> 00:29:07.330
but crystalline silicon
technology was going faster

00:29:07.330 --> 00:29:08.640
than the others.

00:29:08.640 --> 00:29:10.640
And in part, this
was due to the fact

00:29:10.640 --> 00:29:14.200
that crystalline silicon
technology was a bit cheaper

00:29:14.200 --> 00:29:17.150
than thin films at
the time, largely

00:29:17.150 --> 00:29:18.730
driven by the
efficiency parameter

00:29:18.730 --> 00:29:21.080
that we've just seen
in the previous slides.

00:29:21.080 --> 00:29:24.580
And turnkey equipment
was available.

00:29:24.580 --> 00:29:27.880
So if you had capital
someplace in the world,

00:29:27.880 --> 00:29:31.350
anywhere in the world, and the
labor and commodities were such

00:29:31.350 --> 00:29:32.930
that you could
compete in the market,

00:29:32.930 --> 00:29:37.900
and the shipping costs weren't
extremely prohibitive to get

00:29:37.900 --> 00:29:39.870
your product to the
most interesting markets

00:29:39.870 --> 00:29:43.070
in the world, you could
compete because you

00:29:43.070 --> 00:29:44.344
could buy turnkey equipment.

00:29:44.344 --> 00:29:45.760
Even if you knew
nothing about how

00:29:45.760 --> 00:29:47.780
to manufacture a solar
cell, knew nothing

00:29:47.780 --> 00:29:49.804
about semiconductors
or solid-state physics,

00:29:49.804 --> 00:29:51.470
what we've discussed
here in this class,

00:29:51.470 --> 00:29:54.260
you could still go out and buy
a turnkey manufacturing line

00:29:54.260 --> 00:29:56.110
and get technicians to
come in and teach you

00:29:56.110 --> 00:29:58.610
how to manufacture solar cells.

00:29:58.610 --> 00:30:01.540
That was the beauty of
these equipment companies.

00:30:01.540 --> 00:30:03.860
And so the equipment
companies specialized

00:30:03.860 --> 00:30:05.710
in crystalline
silicon technologies.

00:30:05.710 --> 00:30:08.550
And as such, many
of the turnkey lines

00:30:08.550 --> 00:30:13.960
that grew up out of
Greenfield factories,

00:30:13.960 --> 00:30:18.260
especially in China, Taiwan, and
other places around the world,

00:30:18.260 --> 00:30:20.490
leveraged these
turnkey manufacturers

00:30:20.490 --> 00:30:21.240
to a great extent.

00:30:25.710 --> 00:30:27.350
So price, markets,
and subsidies.

00:30:27.350 --> 00:30:31.670
We're going to be
looking at-- it's

00:30:31.670 --> 00:30:33.440
a bit of a hodgepodge
in the sense,

00:30:33.440 --> 00:30:37.220
if we're addressing cost, price,
and manufacturing all together

00:30:37.220 --> 00:30:40.370
in one big stew, I
think that's useful

00:30:40.370 --> 00:30:43.260
because the three
are interrelated.

00:30:43.260 --> 00:30:46.120
A price is difficult
to set without a cost,

00:30:46.120 --> 00:30:49.080
and the manufacturing
is part of that story

00:30:49.080 --> 00:30:51.280
and can help us tease
apart what exactly

00:30:51.280 --> 00:30:54.800
is going on right now with
Solyndra, with SolarWorld,

00:30:54.800 --> 00:30:55.400
and so forth.

00:30:55.400 --> 00:30:56.250
OK.

00:30:56.250 --> 00:30:58.890
So let's start with
customer needs.

00:30:58.890 --> 00:31:01.350
Just to acquaint ourselves,
if we're talking about price,

00:31:01.350 --> 00:31:05.820
we can't divorce price
from our customer.

00:31:05.820 --> 00:31:08.140
And in terms of what
our customer needs are,

00:31:08.140 --> 00:31:09.540
we have on-grid applications.

00:31:09.540 --> 00:31:11.490
That's represented
by, for example,

00:31:11.490 --> 00:31:13.789
a solar system on my house,
on the student center.

00:31:13.789 --> 00:31:15.580
These are systems that
are tied to the grid

00:31:15.580 --> 00:31:19.980
and using the grid as a battery
to store the excess energy.

00:31:19.980 --> 00:31:22.880
Off-grid, this represents, for
example, the Lighting Africa

00:31:22.880 --> 00:31:26.110
project here within our class.

00:31:26.110 --> 00:31:28.870
These are folks who don't have
access to an electrical grid

00:31:28.870 --> 00:31:32.255
and who need to have
that electricity there.

00:31:32.255 --> 00:31:33.880
So while people on
the grid are worried

00:31:33.880 --> 00:31:36.580
about cents per kilowatt
hour, people off the grid

00:31:36.580 --> 00:31:38.920
might be worried about the
dollars per hour of light.

00:31:38.920 --> 00:31:41.472
That might be their
metric of merit.

00:31:41.472 --> 00:31:42.930
So from the customer's
perspective,

00:31:42.930 --> 00:31:44.300
what is their value?

00:31:44.300 --> 00:31:48.407
What do they get from the
solar PV system on the roof?

00:31:48.407 --> 00:31:50.240
These parameters under
here, underneath each

00:31:50.240 --> 00:31:53.040
of the pictures, represent
the value parameter.

00:31:53.040 --> 00:31:55.700
And again, it's a
very cartoonish way

00:31:55.700 --> 00:31:56.699
of thinking about it.

00:31:56.699 --> 00:31:59.240
It's a lot more complex when
you start getting into the weeds

00:31:59.240 --> 00:32:01.239
and figure out what the
customer actually wants.

00:32:01.239 --> 00:32:05.030
Reliability factors in,
access to the product, repair,

00:32:05.030 --> 00:32:06.910
reliability, and so
forth, factor in as well.

00:32:06.910 --> 00:32:09.420
There are many factors
that add in to value.

00:32:09.420 --> 00:32:12.290
But what we've done
here is emphasize

00:32:12.290 --> 00:32:14.535
the biggest levers,
if you will, and some

00:32:14.535 --> 00:32:16.910
of the biggest differences
between different applications

00:32:16.910 --> 00:32:18.380
of PV.

00:32:18.380 --> 00:32:21.780
If you're looking at the solar
panels on top of the Toyota

00:32:21.780 --> 00:32:24.996
Prius, you might not
really care about the cents

00:32:24.996 --> 00:32:26.370
per kilowatt hour,
because you're

00:32:26.370 --> 00:32:29.880
paying $20,000 for your car, so
what's a few extra dollars here

00:32:29.880 --> 00:32:30.590
or there?

00:32:30.590 --> 00:32:32.120
But you might care
about the watts

00:32:32.120 --> 00:32:34.810
per meter squared, which is
really an efficiency parameter.

00:32:34.810 --> 00:32:37.302
You might want it to
satisfy a certain function

00:32:37.302 --> 00:32:39.260
that it could not do
otherwise if the panel was

00:32:39.260 --> 00:32:40.614
too low efficiency.

00:32:40.614 --> 00:32:42.780
If you're sending something
into outer space, again,

00:32:42.780 --> 00:32:44.738
you might not care about
the manufacturing cost

00:32:44.738 --> 00:32:47.640
of the panel, , because the
shipment costs, in other words,

00:32:47.640 --> 00:32:50.139
putting it onto the rocket and
sending it up in outer space,

00:32:50.139 --> 00:32:51.810
is $10,000 per kilogram.

00:32:51.810 --> 00:32:54.150
You might be more worried
about, how much does it weigh?

00:32:54.150 --> 00:32:57.020
What is the specific
power of the solar panels?

00:32:57.020 --> 00:32:59.730
Likewise, if you're installing
them in a big-box company,

00:32:59.730 --> 00:33:01.960
for example at Walmart
or Kmart, and you

00:33:01.960 --> 00:33:04.720
have a big flat roof
that isn't very strong,

00:33:04.720 --> 00:33:07.730
isn't well reinforced, you
might care about that parameter

00:33:07.730 --> 00:33:10.570
as well, or the grams
per meter squared.

00:33:10.570 --> 00:33:12.500
I'm sure that
factors in somewhere.

00:33:12.500 --> 00:33:13.470
Actually, it doesn't.

00:33:13.470 --> 00:33:16.690
So there are many
other parameters

00:33:16.690 --> 00:33:21.070
that could matter for a
particular application.

00:33:21.070 --> 00:33:23.880
Dollars per meter squared
for aesthetics for a building

00:33:23.880 --> 00:33:25.700
integrated system.

00:33:25.700 --> 00:33:27.940
I've had architects
come to me and say,

00:33:27.940 --> 00:33:30.119
can you make a
yellow solar panel?

00:33:30.119 --> 00:33:31.660
I'd really like a
yellow solar panel.

00:33:31.660 --> 00:33:33.460
And here I am thinking
of the solar spectrum

00:33:33.460 --> 00:33:35.959
and seeing the biggest, the
peak of the solar spectrum being

00:33:35.959 --> 00:33:38.700
reflected away from the panel
into the observer's eye,

00:33:38.700 --> 00:33:41.480
and I'm thinking, well, I can
make it kind of dark yellow,

00:33:41.480 --> 00:33:42.720
yellowish.

00:33:42.720 --> 00:33:43.220
OK.

00:33:43.220 --> 00:33:45.360
How much are you willing to pay?

00:33:45.360 --> 00:33:48.580
And there are many different
parameters here that matter

00:33:48.580 --> 00:33:50.820
for the customer.

00:33:50.820 --> 00:33:53.345
This one right here is
a concentrator system,

00:33:53.345 --> 00:33:55.470
so you have these optics
that concentrate the light

00:33:55.470 --> 00:33:57.450
into a small little
spot, and they're

00:33:57.450 --> 00:34:00.230
looking at the watts
per millimeter squared.

00:34:00.230 --> 00:34:03.010
A very small device, how
much power does it output?

00:34:03.010 --> 00:34:06.330
The cost of the actual
device is almost irrelevant.

00:34:06.330 --> 00:34:08.050
It's the power
output that matters,

00:34:08.050 --> 00:34:11.070
because the majority of the cost
is sunk in all the commodity

00:34:11.070 --> 00:34:12.290
materials around it.

00:34:12.290 --> 00:34:12.790
All right.

00:34:12.790 --> 00:34:15.159
So we're getting into
this because we're

00:34:15.159 --> 00:34:17.710
going to focus largely
on on-grid applications,

00:34:17.710 --> 00:34:19.392
and I'm going to
explain to you why.

00:34:19.392 --> 00:34:21.100
If you're in the
Lighting Africa project,

00:34:21.100 --> 00:34:23.230
you probably care
mostly about that.

00:34:23.230 --> 00:34:24.909
But the on-grid
applications currently

00:34:24.909 --> 00:34:27.860
comprise 95% of
the current market.

00:34:27.860 --> 00:34:28.850
Substitution economics.

00:34:28.850 --> 00:34:32.310
We're substituting
PV with, or we're

00:34:32.310 --> 00:34:34.850
substituting fossil
fuel-based power,

00:34:34.850 --> 00:34:38.130
on average in the United States,
somewhere around 600 grams

00:34:38.130 --> 00:34:40.510
of CO2 per kilowatt hour.

00:34:40.510 --> 00:34:42.219
We're substituting
that with PV which

00:34:42.219 --> 00:34:46.580
is on the order of a
factor of 5 to 20 lower.

00:34:46.580 --> 00:34:50.044
So what types of grid
electricity will PV substitute?

00:34:50.044 --> 00:34:51.960
What does this mean for
traditional generation

00:34:51.960 --> 00:34:53.909
companies, also called gencos?

00:34:53.909 --> 00:34:57.365
And what is a fair selling
price for PV electricity?

00:34:57.365 --> 00:34:58.490
Very interesting questions.

00:34:58.490 --> 00:35:01.290
So in terms of
markets, this is just

00:35:01.290 --> 00:35:05.350
a breakdown of off-grid consumer
applications and on-grid.

00:35:05.350 --> 00:35:08.680
When I first got into this,
somewhere around here,

00:35:08.680 --> 00:35:12.420
it was broken down about
50-50, PV on-grid and off-grid.

00:35:14.990 --> 00:35:18.790
Today, much further down,
off-grid applications, probably

00:35:18.790 --> 00:35:19.882
below 10%.

00:35:19.882 --> 00:35:21.590
I don't know the
precise numbers for you,

00:35:21.590 --> 00:35:24.960
but it's stayed relatively
flat compared to the growth

00:35:24.960 --> 00:35:26.680
rate of the on-grid.

00:35:26.680 --> 00:35:30.150
And the reasons for this, we'll
see in a couple of slides.

00:35:30.150 --> 00:35:35.300
This is the value of PV
electricity, per a 2008 report.

00:35:35.300 --> 00:35:41.370
If you look at what is easily
monetized-- easily monetized

00:35:41.370 --> 00:35:44.650
depending on the policy
in a particular case--

00:35:44.650 --> 00:35:48.550
you have the cost of
fuel, the cost of capital,

00:35:48.550 --> 00:35:53.120
typically for a power plant,
the CO2 emissions offset.

00:35:53.120 --> 00:35:55.060
This is if you have
a price of carbon.

00:35:55.060 --> 00:35:58.820
Notice how small it actually
is, leading a lot of people

00:35:58.820 --> 00:36:02.080
to conclude that maybe carbon
pricing isn't the biggest lever

00:36:02.080 --> 00:36:03.447
for bringing PV onto the grid.

00:36:03.447 --> 00:36:04.780
That's a whole other discussion.

00:36:04.780 --> 00:36:06.280
We can have that later.

00:36:06.280 --> 00:36:07.220
Grid losses.

00:36:07.220 --> 00:36:09.440
These are the transmission
and distribution losses,

00:36:09.440 --> 00:36:12.160
the difference between having
the PV mounted or distributed,

00:36:12.160 --> 00:36:13.970
the power generation
source mounted right

00:36:13.970 --> 00:36:16.736
on the site of use, as opposed
to a centralized location that

00:36:16.736 --> 00:36:18.940
has to distribute it.

00:36:18.940 --> 00:36:21.820
These are much more
difficult to monetize.

00:36:21.820 --> 00:36:24.750
And you can add in here health
impacts related to emissions,

00:36:24.750 --> 00:36:26.720
and so forth.

00:36:26.720 --> 00:36:29.760
You're looking at the
security, the reduced

00:36:29.760 --> 00:36:33.170
risk of having a product
producing electricity

00:36:33.170 --> 00:36:36.280
at a certain known
price for 20 years.

00:36:36.280 --> 00:36:38.820
You're also looking
at tax bonuses.

00:36:38.820 --> 00:36:43.140
These are tricky because
tax policy can change.

00:36:43.140 --> 00:36:46.720
You're looking at uncertainty
in your raw feedstock

00:36:46.720 --> 00:36:49.650
material, the fuel.

00:36:49.650 --> 00:36:52.310
And the need for
backup is something

00:36:52.310 --> 00:36:54.870
that reduces the value
of PV electricity.

00:36:54.870 --> 00:36:56.520
If you, at some
point in the future,

00:36:56.520 --> 00:36:59.660
will need to back up PV
power with something that

00:36:59.660 --> 00:37:01.870
is more dispatchable
and it's still

00:37:01.870 --> 00:37:03.810
an uncertainty in
the power grid,

00:37:03.810 --> 00:37:07.070
that's something that
is reducing the value

00:37:07.070 --> 00:37:08.450
of installing the PV today.

00:37:08.450 --> 00:37:11.330
Some might argue that that
$0.01 negative is actually

00:37:11.330 --> 00:37:14.710
a drastic underestimation
of the risk associated

00:37:14.710 --> 00:37:16.450
with backup power needs.

00:37:16.450 --> 00:37:23.070
But this is a fair look at
what the so-called true value

00:37:23.070 --> 00:37:24.930
of PV electricity
might be, which

00:37:24.930 --> 00:37:26.230
would lead us into pricing.

00:37:26.230 --> 00:37:29.010
If we know the value,
we can enter pricing.

00:37:29.010 --> 00:37:31.460
And not many people can
argue with this, because this

00:37:31.460 --> 00:37:32.546
is substitution economics.

00:37:32.546 --> 00:37:33.920
This is just
saying, OK, how much

00:37:33.920 --> 00:37:38.660
does it cost to produce
fossil fuel electricity?

00:37:38.660 --> 00:37:41.020
And let's substitute that.

00:37:41.020 --> 00:37:44.110
Anything over here in the yellow
is what people might say, well

00:37:44.110 --> 00:37:47.700
gee, if we really look at the
true price, or the true cost

00:37:47.700 --> 00:37:50.000
of fossil fuels, this
is what it really

00:37:50.000 --> 00:37:51.960
costs if we take away
all of the subsidies

00:37:51.960 --> 00:37:54.390
and add in all of
the externalities.

00:37:54.390 --> 00:37:56.465
What's an externality?

00:37:56.465 --> 00:37:57.590
What's an externality mean?

00:38:00.364 --> 00:38:00.864
Yeah?

00:38:00.864 --> 00:38:02.780
AUDIENCE: Is it
when the consumer

00:38:02.780 --> 00:38:05.954
isn't aware of something that
the [INAUDIBLE] is doing?

00:38:05.954 --> 00:38:06.620
PROFESSOR: Yeah.

00:38:06.620 --> 00:38:09.740
So when the consumer--
or let's put it this way.

00:38:09.740 --> 00:38:13.770
When the true
impact, price impact,

00:38:13.770 --> 00:38:16.270
is not factored into
the selling price.

00:38:16.270 --> 00:38:20.370
Let's say-- hm.

00:38:20.370 --> 00:38:24.510
Let's say that I sell
you a miracle drug.

00:38:24.510 --> 00:38:27.810
And it allows you to be
5 times more productive

00:38:27.810 --> 00:38:29.410
than you are right now.

00:38:29.410 --> 00:38:34.740
But then every time
you have your needs,

00:38:34.740 --> 00:38:36.840
down the toilet goes
a bunch of chemicals

00:38:36.840 --> 00:38:39.980
that the water plant
now needs to filter out,

00:38:39.980 --> 00:38:41.820
and the water plant
begins failing

00:38:41.820 --> 00:38:45.930
some of its standard tests when
they measure water quality.

00:38:45.930 --> 00:38:48.601
So they add in some more
filters into their system.

00:38:48.601 --> 00:38:50.600
They figure out how to
get rid of this compound.

00:38:50.600 --> 00:38:53.252
And now all of the water
treatment facilities

00:38:53.252 --> 00:38:55.460
around the country begin
adding in these new filters,

00:38:55.460 --> 00:38:57.543
and it costs somewhere on
the order of $1 billion.

00:38:57.543 --> 00:38:58.730
That is an externality.

00:38:58.730 --> 00:38:59.950
That's something
that wasn't factored

00:38:59.950 --> 00:39:01.880
into the price of selling
you that miracle drug that

00:39:01.880 --> 00:39:03.830
allowed you to be 5
times more productive.

00:39:03.830 --> 00:39:07.000
It's an example of
how everybody pays

00:39:07.000 --> 00:39:11.300
for the acts or the
purchases of a few.

00:39:11.300 --> 00:39:14.640
And that's the case with
energy production as well.

00:39:17.430 --> 00:39:21.330
So it's difficult to
argue in economic terms--

00:39:21.330 --> 00:39:24.020
or it's difficult to
put a specific price

00:39:24.020 --> 00:39:26.090
on externalities, because
these could be things

00:39:26.090 --> 00:39:27.850
like premature deaths.

00:39:27.850 --> 00:39:30.214
And there are statisticians
who will calculate

00:39:30.214 --> 00:39:31.880
the number of people
who die prematurely

00:39:31.880 --> 00:39:35.676
as a result of exposure
to mercury or cadmium

00:39:35.676 --> 00:39:37.050
or some other
emission coming off

00:39:37.050 --> 00:39:39.130
of a fossil fuel-burning plant.

00:39:39.130 --> 00:39:41.500
But then how do
you monetize that?

00:39:41.500 --> 00:39:44.230
You have to assign a
value to a human life

00:39:44.230 --> 00:39:48.366
and say that a certain economic
value was lost, both in terms

00:39:48.366 --> 00:39:49.990
of productivity cost,
but also in terms

00:39:49.990 --> 00:39:53.110
of the investment due
to that person dying.

00:39:53.110 --> 00:39:54.560
And they do that.

00:39:54.560 --> 00:39:57.580
In government, there is
a value to a human life,

00:39:57.580 --> 00:40:01.410
and it can vary somewhat
from one group to another.

00:40:01.410 --> 00:40:05.710
But it's very difficult
to price in externalities.

00:40:05.710 --> 00:40:08.450
And that's why oftentimes
these things in yellow

00:40:08.450 --> 00:40:10.780
here are neglected
or not considered.

00:40:10.780 --> 00:40:12.600
For the purposes of
today's discussion,

00:40:12.600 --> 00:40:14.950
we're just going to factor
in mostly these parameters

00:40:14.950 --> 00:40:19.680
right here because we have
a greater hold on them.

00:40:19.680 --> 00:40:22.820
So in terms of PV
installations worldwide

00:40:22.820 --> 00:40:25.100
cumulative, I want to
compare and contrast

00:40:25.100 --> 00:40:29.670
where our customers are versus
where our manufacturing is.

00:40:29.670 --> 00:40:32.580
And we'll get to some
really interesting questions

00:40:32.580 --> 00:40:33.750
associated with that.

00:40:33.750 --> 00:40:36.770
So look at that.

00:40:36.770 --> 00:40:38.549
Where would you expect
PV to be installed?

00:40:38.549 --> 00:40:40.840
You'd expect it to be installed
in the sunniest places.

00:40:40.840 --> 00:40:41.389
Why?

00:40:41.389 --> 00:40:42.930
Because the amount
of energy produced

00:40:42.930 --> 00:40:47.190
is related, proportional, to
the amount of solar resource

00:40:47.190 --> 00:40:49.280
that's available in that spot.

00:40:49.280 --> 00:40:51.970
But from a customer's
point of view,

00:40:51.970 --> 00:40:53.400
solar resource isn't everything.

00:40:53.400 --> 00:40:56.250
They're also looking at
the cost of displacement.

00:40:56.250 --> 00:40:58.622
What is the price of
electricity I'm paying?

00:40:58.622 --> 00:41:00.080
And how much of
that can I displace

00:41:00.080 --> 00:41:01.940
with my PV electricity?

00:41:01.940 --> 00:41:05.800
So there's some more
sophisticated concern here.

00:41:05.800 --> 00:41:08.250
And further, the
price of electricity

00:41:08.250 --> 00:41:11.910
might not only be dictated by
the true cost of production

00:41:11.910 --> 00:41:14.224
of the fossil fuel power plants.

00:41:14.224 --> 00:41:15.890
There may be a few
governments out there

00:41:15.890 --> 00:41:17.920
that say, well, look at
all these externalities.

00:41:17.920 --> 00:41:19.920
We want to begin
factoring those in.

00:41:19.920 --> 00:41:22.030
We want to provide an
incentive for people

00:41:22.030 --> 00:41:24.490
to produce PV electricity.

00:41:24.490 --> 00:41:26.430
That's commonly referred
to as a subsidy.

00:41:26.430 --> 00:41:28.950
So if we look at the
total installation,

00:41:28.950 --> 00:41:31.540
EU is really leading
the charge here.

00:41:31.540 --> 00:41:33.370
And we saw on those
installation maps,

00:41:33.370 --> 00:41:36.340
there's not a heck of a
lot of sun there in the EU.

00:41:36.340 --> 00:41:41.880
Germany has a solar radiance
similar to what we have here

00:41:41.880 --> 00:41:43.310
in the northeast of the country.

00:41:43.310 --> 00:41:46.030
And we know that most of the PV
installed in the United States

00:41:46.030 --> 00:41:48.000
is going into the southwest.

00:41:48.000 --> 00:41:53.650
Japan, rest of world, USA,
a relatively small fraction

00:41:53.650 --> 00:41:55.370
up there in China.

00:41:55.370 --> 00:41:56.450
Tiny, tiny little bleep.

00:41:56.450 --> 00:41:59.550
China's the peach one
right up there at the top.

00:41:59.550 --> 00:42:04.530
So I would say less than 5%
of total installed worldwide.

00:42:04.530 --> 00:42:06.360
Actually this is 1.5%.

00:42:06.360 --> 00:42:08.920
You can read it off
right here, the division.

00:42:08.920 --> 00:42:09.420
OK.

00:42:09.420 --> 00:42:13.850
So we have a very
interesting perspective

00:42:13.850 --> 00:42:16.680
here about our customers.

00:42:16.680 --> 00:42:19.970
This is a breakdown, again,
in terms of customers again.

00:42:19.970 --> 00:42:22.530
A little bit better detail.

00:42:22.530 --> 00:42:26.000
Instead of just seeing EU,
we're looking at a variety

00:42:26.000 --> 00:42:28.510
of different customers.

00:42:28.510 --> 00:42:30.370
And this is new
installations worldwide.

00:42:30.370 --> 00:42:32.537
So if, for example,
this blue bar goes down,

00:42:32.537 --> 00:42:34.370
it just means we installed
less PV this year

00:42:34.370 --> 00:42:35.320
than we did last year.

00:42:35.320 --> 00:42:37.819
It doesn't mean that the total
amount on the grid went down.

00:42:37.819 --> 00:42:39.420
It's just new PV installs.

00:42:39.420 --> 00:42:42.420
So the blue down
here is now Germany.

00:42:42.420 --> 00:42:45.670
We have these countries that
are like flashes in the pan.

00:42:45.670 --> 00:42:48.260
Italy, for example--
sorry, Spain.

00:42:48.260 --> 00:42:51.570
This one, Spain, grew up really
quick and then disappeared.

00:42:51.570 --> 00:42:55.280
Italy grew up really quick
and then is shrinking.

00:42:55.280 --> 00:42:57.560
If we extended out to
2011, it would be back

00:42:57.560 --> 00:42:59.320
to a very small amount.

00:42:59.320 --> 00:43:01.490
We have USA that's
consistently growing.

00:43:01.490 --> 00:43:03.150
That's a nice healthy market.

00:43:03.150 --> 00:43:06.540
In Germany, that's
growing as well.

00:43:06.540 --> 00:43:07.810
Wow.

00:43:07.810 --> 00:43:09.420
What an interesting dynamic.

00:43:09.420 --> 00:43:12.500
Why did Spain have
this flash in the pan

00:43:12.500 --> 00:43:14.220
and then shrink suddenly?

00:43:14.220 --> 00:43:17.500
What did they do differently
than Germany did?

00:43:17.500 --> 00:43:19.190
We'll start asking
those questions

00:43:19.190 --> 00:43:21.484
and answering them
in a few slides.

00:43:21.484 --> 00:43:23.400
If you really want to
get detailed information

00:43:23.400 --> 00:43:25.870
about the US, where PV is
installed in the United

00:43:25.870 --> 00:43:28.370
States, one of
your MIT colleagues

00:43:28.370 --> 00:43:32.240
and a bunch of NREL folks
got together and put up

00:43:32.240 --> 00:43:37.740
this beautiful archive, if
you will, compendex of as many

00:43:37.740 --> 00:43:39.150
PV installations
as they possibly

00:43:39.150 --> 00:43:40.510
could in the United States.

00:43:40.510 --> 00:43:41.730
So that's the website.

00:43:41.730 --> 00:43:44.490
Unfortunately, I didn't
include it in your slides,

00:43:44.490 --> 00:43:46.520
so you might want to
write that down if you're

00:43:46.520 --> 00:43:48.830
curious about it.

00:43:48.830 --> 00:43:50.060
Wonderful resource.

00:43:50.060 --> 00:43:53.090
Again, NREL, National
Renewable Energy Laboratory

00:43:53.090 --> 00:43:54.960
based in Golden, Colorado.

00:43:54.960 --> 00:43:57.342
And you can see the
install distribution

00:43:57.342 --> 00:43:58.550
throughout the United States.

00:43:58.550 --> 00:44:03.550
Again, a bias toward states that
have high electricity prices,

00:44:03.550 --> 00:44:08.190
like New York and Massachusetts,
and lots of sun, like Arizona.

00:44:08.190 --> 00:44:11.350
The price of electricity
is rather low in Arizona.

00:44:11.350 --> 00:44:12.920
It's below $0.10
per kilowatt hour,

00:44:12.920 --> 00:44:16.560
but the amount of sun
available is very high.

00:44:16.560 --> 00:44:18.800
Whereas California has both.

00:44:18.800 --> 00:44:19.330
OK.

00:44:19.330 --> 00:44:21.060
So we talked about the customer.

00:44:21.060 --> 00:44:23.850
Now here we're talking
about the manufacturers.

00:44:23.850 --> 00:44:25.860
So we have the
customers, the demand;

00:44:25.860 --> 00:44:28.900
the manufacturers, the supply.

00:44:28.900 --> 00:44:32.360
Obviously, the manufacturers
are growing at the same rate

00:44:32.360 --> 00:44:33.710
as our customers are installing.

00:44:33.710 --> 00:44:35.740
Even faster, mind you.

00:44:35.740 --> 00:44:39.160
Inventory rates are almost,
somewhere between 25% and 50%

00:44:39.160 --> 00:44:39.660
nowadays.

00:44:39.660 --> 00:44:43.730
So the production has grown
faster than the demand has,

00:44:43.730 --> 00:44:45.780
at least at current prices.

00:44:45.780 --> 00:44:48.370
What we've done right here,
or what is done right here

00:44:48.370 --> 00:44:52.930
is a normalization
by market share,

00:44:52.930 --> 00:44:55.340
just to demonstrate how
the market dynamics are

00:44:55.340 --> 00:44:57.150
changing as the industry grows.

00:44:57.150 --> 00:44:59.251
So from 2005 to 2009,
this doesn't mean

00:44:59.251 --> 00:45:00.500
that the industry stayed flat.

00:45:00.500 --> 00:45:02.330
It continued growing
at a breakneck pace,

00:45:02.330 --> 00:45:06.870
but the countries which
comprise the manufacturers

00:45:06.870 --> 00:45:08.364
have changed significantly.

00:45:08.364 --> 00:45:09.780
So let's look at
Europe first off.

00:45:09.780 --> 00:45:12.330
Europe pretty much
started really feeling it

00:45:12.330 --> 00:45:13.960
during the financial crisis.

00:45:13.960 --> 00:45:19.270
They couldn't keep
expanding in 2008,

00:45:19.270 --> 00:45:22.610
and then when prices really
started to drop, 2009,

00:45:22.610 --> 00:45:26.050
as we saw, they didn't
continue expanding.

00:45:26.050 --> 00:45:28.690
Were a bit uncompetitive.

00:45:28.690 --> 00:45:33.580
US, shown here in the
green, again, dropping.

00:45:33.580 --> 00:45:37.960
India, I guess sort
of growing now.

00:45:37.960 --> 00:45:40.440
It's definitely on the upswing.

00:45:40.440 --> 00:45:43.920
Japan decreased
considerably, considerably.

00:45:43.920 --> 00:45:47.400
During the 1990s
and the early 2000s,

00:45:47.400 --> 00:45:51.130
Japan had the largest solar
company in the world, Sharp,

00:45:51.130 --> 00:45:52.830
better known for
microelectronics

00:45:52.830 --> 00:45:54.660
that you might find
around your house.

00:45:54.660 --> 00:45:57.560
Somehow, some way, the
executives at Sharp

00:45:57.560 --> 00:45:59.030
saw this coming.

00:45:59.030 --> 00:46:02.960
Saw a huge rise in demand coming
from Europe, and ramped up

00:46:02.960 --> 00:46:04.545
capacity in Japan.

00:46:04.545 --> 00:46:05.920
And when they did
that, they were

00:46:05.920 --> 00:46:07.544
able to address large
portions of that.

00:46:07.544 --> 00:46:09.620
They made a healthy
profit for several years,

00:46:09.620 --> 00:46:13.770
and then they just
stopped expanding in PV.

00:46:13.770 --> 00:46:16.441
Part of it might have been
they saw the market dynamics

00:46:16.441 --> 00:46:16.940
changing.

00:46:16.940 --> 00:46:20.230
They saw their costs,
their manufacturing costs,

00:46:20.230 --> 00:46:22.650
relative to, for
example, China--

00:46:22.650 --> 00:46:27.590
this is the orange right
here-- and Taiwan, above China,

00:46:27.590 --> 00:46:31.850
above the orange-- and comprised
today-- this is 2009 numbers,

00:46:31.850 --> 00:46:33.890
if we fast-forward
to today in 2011,

00:46:33.890 --> 00:46:38.490
China and Taiwan comprise
55% to 60% of the PV industry

00:46:38.490 --> 00:46:39.080
worldwide.

00:46:39.080 --> 00:46:43.072
Production,
manufacturing, production.

00:46:43.072 --> 00:46:44.030
So what does that mean?

00:46:44.030 --> 00:46:49.060
Let's explore together some
of the interesting results

00:46:49.060 --> 00:46:51.170
of that.

00:46:51.170 --> 00:46:52.880
Let's look at some
of the good things.

00:46:52.880 --> 00:46:56.150
What are some of the good
things about production going

00:46:56.150 --> 00:47:00.820
to China and Taiwan
from, say, Europe

00:47:00.820 --> 00:47:03.010
and the United States,
of solar panels?

00:47:03.010 --> 00:47:05.020
Let's look at that for a second.

00:47:05.020 --> 00:47:07.850
So we'll look at the glass
half full perspective.

00:47:07.850 --> 00:47:09.442
What are some of
the good things?

00:47:09.442 --> 00:47:11.858
AUDIENCE: It's more likely
that these developing countries

00:47:11.858 --> 00:47:12.934
will use solar panels?

00:47:12.934 --> 00:47:15.100
PROFESSOR: More likely that
the developing countries

00:47:15.100 --> 00:47:17.490
will use solar panels?

00:47:17.490 --> 00:47:20.040
Was that the case over here?

00:47:20.040 --> 00:47:21.550
Not really.

00:47:21.550 --> 00:47:22.650
It's going.

00:47:22.650 --> 00:47:23.730
It's going.

00:47:23.730 --> 00:47:25.350
OK.

00:47:25.350 --> 00:47:28.030
I suppose, if the
technology's available there.

00:47:28.030 --> 00:47:30.940
Maybe the technology isn't
quite cost competitive

00:47:30.940 --> 00:47:34.200
with local electricity
yet, so there's

00:47:34.200 --> 00:47:36.660
isn't that demand pull locally.

00:47:36.660 --> 00:47:39.560
Maybe the realization
is that, well, goodness,

00:47:39.560 --> 00:47:41.110
we can address this
European market.

00:47:41.110 --> 00:47:43.110
They're willing to pay a
lot more for the panels

00:47:43.110 --> 00:47:45.414
than we are, so we might
as well export right now.

00:47:45.414 --> 00:47:46.830
But at some point
in the future we

00:47:46.830 --> 00:47:48.650
might be able to
satisfy internal demand.

00:47:48.650 --> 00:47:49.190
Yeah.

00:47:49.190 --> 00:47:52.420
And there are gigawatt plants,
PV installs going up in China

00:47:52.420 --> 00:47:53.790
right now.

00:47:53.790 --> 00:47:56.720
In part due to increased
electricity demand, in part

00:47:56.720 --> 00:47:58.960
due to weakened demand
elsewhere in the world,

00:47:58.960 --> 00:48:01.410
and a very large
manufacturing base in China

00:48:01.410 --> 00:48:03.920
that has to put their
panels out somewhere.

00:48:03.920 --> 00:48:06.660
So there is, yes, adoption.

00:48:06.660 --> 00:48:08.810
I'll cede that point.

00:48:08.810 --> 00:48:09.608
Yeah?

00:48:09.608 --> 00:48:11.887
AUDIENCE: Lower cost
to the consumer?

00:48:11.887 --> 00:48:13.720
PROFESSOR: Lower cost
to the consumer, yeah.

00:48:13.720 --> 00:48:17.400
So in the United States, if you
look at the price of installing

00:48:17.400 --> 00:48:20.210
a PV system in
the United States,

00:48:20.210 --> 00:48:24.660
the price of installing the
PV system is around $5.20

00:48:24.660 --> 00:48:26.590
on your roof today.

00:48:26.590 --> 00:48:32.030
The cost of buying the module
is around $1.03, $1.05,

00:48:32.030 --> 00:48:33.160
from China.

00:48:33.160 --> 00:48:35.710
And so that means that
80% of the profit margin

00:48:35.710 --> 00:48:40.750
right now is being gobbled
up by a US company.

00:48:40.750 --> 00:48:42.730
The price came down--
let's see, when

00:48:42.730 --> 00:48:45.070
I installed the
panels on my roof,

00:48:45.070 --> 00:48:46.670
I got a little bit
of a better deal.

00:48:46.670 --> 00:48:51.220
But I would say the average
price for a PV system in 2007

00:48:51.220 --> 00:48:54.650
must've been somewhere
around $8 per watt-peak.

00:48:54.650 --> 00:48:57.370
And now the price is
down at around $5.20.

00:48:57.370 --> 00:48:59.860
That said, in
Europe, in Germany,

00:48:59.860 --> 00:49:03.790
which has 10 times more
installed PV than the US does,

00:49:03.790 --> 00:49:06.820
the price of installing
a PV system on your roof

00:49:06.820 --> 00:49:09.920
could be below 3
euros per watt-peak.

00:49:09.920 --> 00:49:13.210
So the price is lower
because the profit

00:49:13.210 --> 00:49:16.910
margin that the installers are
getting right now is smaller.

00:49:16.910 --> 00:49:17.990
So yes, absolutely.

00:49:17.990 --> 00:49:20.390
Lower cost product
in US markets.

00:49:20.390 --> 00:49:23.250
That isn't the whole story
about what you're going to pay,

00:49:23.250 --> 00:49:25.000
because there's still
the installer that's

00:49:25.000 --> 00:49:27.680
stuck in between the Chinese
manufacturer of the module

00:49:27.680 --> 00:49:30.390
and you serving as
the middle person.

00:49:30.390 --> 00:49:33.180
Providing you value,
still, but still extracting

00:49:33.180 --> 00:49:36.650
a very large profit right now,
a disproportionately large

00:49:36.650 --> 00:49:39.410
profit, shall I say.

00:49:39.410 --> 00:49:42.074
What are some of the
other good things.

00:49:42.074 --> 00:49:43.070
Yeah?

00:49:43.070 --> 00:49:45.560
AUDIENCE: If cost in
big markets is reduced,

00:49:45.560 --> 00:49:48.060
then that could lead to
more market penetration,

00:49:48.060 --> 00:49:52.021
and then the [INAUDIBLE] adopted
more, and even if prices go up,

00:49:52.021 --> 00:49:53.270
it might maintain [INAUDIBLE].

00:49:53.270 --> 00:49:53.936
PROFESSOR: Yeah.

00:49:53.936 --> 00:49:56.270
And so what you're
leading to here

00:49:56.270 --> 00:49:59.310
is really what the demand
curve of PV looks like.

00:49:59.310 --> 00:50:02.490
And in the past,
if you look at what

00:50:02.490 --> 00:50:04.917
people are willing
to pay for PV,

00:50:04.917 --> 00:50:07.250
let's say-- let's convert it
into dollars per watt-peak,

00:50:07.250 --> 00:50:11.460
since that's the
universal unit of PV cost.

00:50:11.460 --> 00:50:13.760
And this is the, let's
call it total market

00:50:13.760 --> 00:50:22.010
size in terms of
watt-peak, and this being

00:50:22.010 --> 00:50:23.780
a very, very large number.

00:50:23.780 --> 00:50:25.460
What you can do
to-- first order.

00:50:25.460 --> 00:50:28.010
Just how would we
construct the demand curve

00:50:28.010 --> 00:50:29.730
for the United States for PV.

00:50:29.730 --> 00:50:32.890
How would we go
about doing that?

00:50:32.890 --> 00:50:35.500
How much are people willing
to pay for their PV modules

00:50:35.500 --> 00:50:37.095
to offset the cost
of electricity?

00:50:37.095 --> 00:50:39.095
Well, to do it right, we
need the levelized cost

00:50:39.095 --> 00:50:40.110
of electricity analysis.

00:50:40.110 --> 00:50:42.068
We'd assume we're borrowing
money from the bank

00:50:42.068 --> 00:50:44.920
and do those fancy economics
that involve something

00:50:44.920 --> 00:50:47.830
to the power of something, and
that being the interest rate,

00:50:47.830 --> 00:50:49.390
and then we
calculate it through.

00:50:49.390 --> 00:50:50.340
Just a first order.

00:50:50.340 --> 00:50:51.750
Hand-wavy Mickey-Mousey.

00:50:51.750 --> 00:50:54.540
We might take into consideration
the manufacturing cost

00:50:54.540 --> 00:50:55.661
of the module.

00:50:55.661 --> 00:50:56.160
Sorry.

00:50:56.160 --> 00:50:56.881
Back up one step.

00:50:56.881 --> 00:50:58.630
We might take into
consideration the cents

00:50:58.630 --> 00:51:01.860
per kilowatt hour of the
electricity that we're

00:51:01.860 --> 00:51:06.920
getting from the grid and the
insulation, the solar resource,

00:51:06.920 --> 00:51:09.400
that is available at
that particular location.

00:51:09.400 --> 00:51:11.680
So if we have data
granular to the state

00:51:11.680 --> 00:51:14.279
level of the price
of electricity

00:51:14.279 --> 00:51:16.570
on average throughout the
state, and the solar resource

00:51:16.570 --> 00:51:18.610
availability on average
throughout the state,

00:51:18.610 --> 00:51:21.570
we can immediately comprise
150 markets in the US.

00:51:21.570 --> 00:51:25.500
Residential, commercial,
industrial for 50 states.

00:51:25.500 --> 00:51:27.770
And then we'll have a demand
curve, a very simple one,

00:51:27.770 --> 00:51:29.894
for how much people are
willing to pay for their PV

00:51:29.894 --> 00:51:30.440
electricity.

00:51:30.440 --> 00:51:31.940
And it looks more
or less like this,

00:51:31.940 --> 00:51:36.020
if you start working it out.

00:51:36.020 --> 00:51:37.090
Rough sketch.

00:51:37.090 --> 00:51:38.370
Rough sketch.

00:51:38.370 --> 00:51:40.445
This is Hawaii.

00:51:40.445 --> 00:51:42.320
Those poor critters over
there, although they

00:51:42.320 --> 00:51:44.590
have beautiful sun and
enjoy a wonderful life,

00:51:44.590 --> 00:51:46.260
they're paying a lot
for their energy,

00:51:46.260 --> 00:51:48.010
because they're on a
few rocks out there.

00:51:48.010 --> 00:51:50.310
They have no natural
resource under the ground

00:51:50.310 --> 00:51:53.150
to speak of, except
geothermal, I suppose.

00:51:53.150 --> 00:51:55.374
But they're shipping
in a lot of their fuel.

00:51:55.374 --> 00:51:57.790
That's why if you've ever gone
to Hawaii and rented a car,

00:51:57.790 --> 00:52:00.260
you are surprised at
the sticker shock when

00:52:00.260 --> 00:52:03.170
you go off to the gas
station and try to refuel.

00:52:03.170 --> 00:52:06.840
Their price of electricity is
about $0.30 per kilowatt hour

00:52:06.840 --> 00:52:08.910
residential.

00:52:08.910 --> 00:52:11.040
But it's a very,
very tiny market.

00:52:11.040 --> 00:52:13.320
So you're not going to be
able to satisfy much demand.

00:52:13.320 --> 00:52:15.861
You're not going to be able to
produce too many panels there.

00:52:15.861 --> 00:52:17.775
You start having
interesting things happen

00:52:17.775 --> 00:52:19.150
when you start
hitting the bigger

00:52:19.150 --> 00:52:21.240
markets, like the
tiers four and five

00:52:21.240 --> 00:52:23.717
of California, Texas, New York.

00:52:23.717 --> 00:52:26.050
These are big markets that
have a lot of people in them,

00:52:26.050 --> 00:52:27.740
and they have larger
electricity prices

00:52:27.740 --> 00:52:31.710
and/or large solar
resource available there.

00:52:31.710 --> 00:52:33.970
And so at some point,
the demand curve

00:52:33.970 --> 00:52:37.640
reaches these plateaus, where
if you decrease the price even

00:52:37.640 --> 00:52:40.700
a little bit, of a sudden,
voom, the amount of market

00:52:40.700 --> 00:52:43.280
you can address for
this amount of price

00:52:43.280 --> 00:52:46.620
decrease, the amount of market
that you can address is huge.

00:52:46.620 --> 00:52:49.280
And so the slope of this line,
the slope of the demand curve,

00:52:49.280 --> 00:52:54.000
is indicative of what happens
when you reduce your price just

00:52:54.000 --> 00:52:55.270
a little bit.

00:52:55.270 --> 00:52:57.627
And so yes, producing
cheaper panels,

00:52:57.627 --> 00:52:59.210
when you start hitting
these plateaus,

00:52:59.210 --> 00:53:03.850
can result in massive, massive
demand pull, or market pull.

00:53:03.850 --> 00:53:06.230
So that's another good
reason to have cheap panels.

00:53:06.230 --> 00:53:09.030
Let's look at the flip side.

00:53:09.030 --> 00:53:12.810
What would be some of
the downsides, let's say,

00:53:12.810 --> 00:53:17.020
to module manufacturing
going into China?

00:53:17.020 --> 00:53:19.900
You can assume anything
is on the table.

00:53:19.900 --> 00:53:22.080
CO2 emissions, jobs, et cetera.

00:53:22.080 --> 00:53:24.150
Let's start teasing into
some of those questions

00:53:24.150 --> 00:53:25.192
and looking at some data.

00:53:25.192 --> 00:53:27.316
I might flip back and forth
during the presentation

00:53:27.316 --> 00:53:29.040
if we have to address
specific topics.

00:53:33.694 --> 00:53:37.166
AUDIENCE: Bad politics
if American photovoltaic

00:53:37.166 --> 00:53:42.096
manufacturers fail,
and it reflects poorly

00:53:42.096 --> 00:53:45.547
on the industry and the
American political scene

00:53:45.547 --> 00:53:47.824
and decreases support for that.

00:53:47.824 --> 00:53:48.490
PROFESSOR: Yeah.

00:53:48.490 --> 00:53:50.060
So bad politics.

00:53:50.060 --> 00:53:54.840
In DC, this is often referred to
as the optics of the situation.

00:53:54.840 --> 00:53:58.670
How people observe, or how
people perceive something.

00:53:58.670 --> 00:54:02.820
The litmus test
is not DC itself.

00:54:02.820 --> 00:54:05.370
I was down there on
Thursday, and everybody

00:54:05.370 --> 00:54:08.420
in the solar space was-- I think
it was Wednesday and Thursday,

00:54:08.420 --> 00:54:08.990
yeah.

00:54:08.990 --> 00:54:11.100
Everybody in solar
space was freaking out

00:54:11.100 --> 00:54:14.460
about a particular event that
was going on in Capitol Hill.

00:54:14.460 --> 00:54:16.110
I was telling
folks, don't worry.

00:54:16.110 --> 00:54:18.100
It's not going to--
I mean, in terms

00:54:18.100 --> 00:54:19.285
of the rest of the
country and the perception

00:54:19.285 --> 00:54:21.750
of the rest of the country,
it won't be that significant.

00:54:21.750 --> 00:54:25.600
And I'm sure Secretary Chu will
do a phenomenal job up there

00:54:25.600 --> 00:54:28.230
in front of the
congressional panel.

00:54:28.230 --> 00:54:29.822
He did phenomenal.

00:54:29.822 --> 00:54:31.530
And so people were
really, really worried

00:54:31.530 --> 00:54:35.400
about something that didn't have
too big of an effect outside.

00:54:35.400 --> 00:54:37.460
Granted, what people
should be worried about

00:54:37.460 --> 00:54:39.900
is the impact, for example,
that desequestration

00:54:39.900 --> 00:54:42.350
of the discretionary
funds will have on R&D

00:54:42.350 --> 00:54:45.820
once they start
kicking in in 2013.

00:54:45.820 --> 00:54:52.580
If you mandate a reduced funding
level for funding agencies,

00:54:52.580 --> 00:54:57.620
you will have an impact, and a
long-term one, for that matter.

00:54:57.620 --> 00:55:01.690
So the optics of the situation.

00:55:01.690 --> 00:55:04.330
There has been a slight
decrease in public support,

00:55:04.330 --> 00:55:06.680
I believe on the
order of 5% to 10%

00:55:06.680 --> 00:55:10.570
since the beginning of the
whole Solyndra affair, support

00:55:10.570 --> 00:55:12.210
for solar renewables.

00:55:12.210 --> 00:55:14.310
But the support
remains high, and still

00:55:14.310 --> 00:55:16.870
continues to be high among
Republicans and Democrats

00:55:16.870 --> 00:55:18.320
alike.

00:55:18.320 --> 00:55:20.890
Especially when compared to,
say, congressional approval

00:55:20.890 --> 00:55:23.130
ratings.

00:55:23.130 --> 00:55:25.180
But let's not look down.

00:55:25.180 --> 00:55:27.270
Let's compare
ourselves to higher.

00:55:27.270 --> 00:55:28.170
What else?

00:55:28.170 --> 00:55:29.880
What are some of the
potential negatives?

00:55:29.880 --> 00:55:32.440
Let me start guiding
you, because you

00:55:32.440 --> 00:55:35.540
hear so much about bad
this-- jobs, whatever.

00:55:35.540 --> 00:55:37.720
Let me guide you into
some of the questions

00:55:37.720 --> 00:55:38.970
you might not have thought of.

00:55:41.590 --> 00:55:44.070
There is a colleague
of ours here

00:55:44.070 --> 00:55:46.420
at MIT, Tim
Gutowski's group, that

00:55:46.420 --> 00:55:49.770
is looking into matters
concerning manufacturing of PV

00:55:49.770 --> 00:55:53.700
and installation of PV
from a CO2 perspective.

00:55:53.700 --> 00:55:57.510
So if you manufacture your PV
in a coal-rich country that

00:55:57.510 --> 00:56:00.500
is spewing out
emissions into the air,

00:56:00.500 --> 00:56:04.340
you are going to have a greater
CO2 intensity per kilowatt hour

00:56:04.340 --> 00:56:07.125
of energy that the panels will
produce over their lifetime

00:56:07.125 --> 00:56:08.500
than if the panels
were produced,

00:56:08.500 --> 00:56:13.160
say, in a country that has a
large abundance of hydropower

00:56:13.160 --> 00:56:15.540
or geothermal.

00:56:15.540 --> 00:56:17.690
Or, hey, even
produces solar panels

00:56:17.690 --> 00:56:19.210
using other solar panels.

00:56:19.210 --> 00:56:21.040
That'd be nice.

00:56:21.040 --> 00:56:23.490
But in the growth situation
where solar is rising,

00:56:23.490 --> 00:56:26.070
or the manufacturing capacity
is rising at 60% a year,

00:56:26.070 --> 00:56:28.720
it's a little bit unfeasible
to think that solar will

00:56:28.720 --> 00:56:30.050
power itself forward.

00:56:30.050 --> 00:56:34.020
It almost violates one of
the laws of thermodynamics.

00:56:34.020 --> 00:56:37.260
So we look at the
CO2 balance, right?

00:56:37.260 --> 00:56:40.470
We think about,
gee, what if we were

00:56:40.470 --> 00:56:44.800
to be smart about where we
manufacture and install PV?

00:56:44.800 --> 00:56:47.130
Then it might make more
sense to manufacture PV,

00:56:47.130 --> 00:56:49.830
say, in Norway, which
is based on hydropower,

00:56:49.830 --> 00:56:53.530
and install them in China, which
is largely running on coal.

00:56:53.530 --> 00:56:59.600
So from the CO2, from the global
carbon perspective, again,

00:56:59.600 --> 00:57:03.980
these are externalities
that aren't being priced in.

00:57:03.980 --> 00:57:04.650
OK?

00:57:04.650 --> 00:57:05.460
CO2 and-- yeah.

00:57:05.460 --> 00:57:06.500
Go ahead, please.

00:57:06.500 --> 00:57:08.124
AUDIENCE: But wouldn't
they actually be

00:57:08.124 --> 00:57:12.110
priced in Norway, because
they are in the new emissions

00:57:12.110 --> 00:57:12.760
trading scene?

00:57:12.760 --> 00:57:14.540
I mean, they're not
priced in China,

00:57:14.540 --> 00:57:17.724
but the cost of electricity
in Norway doesn't [INAUDIBLE].

00:57:17.724 --> 00:57:18.390
PROFESSOR: Yeah.

00:57:18.390 --> 00:57:19.960
So point conceded.

00:57:19.960 --> 00:57:22.715
In certain countries,
some of the externalities

00:57:22.715 --> 00:57:23.382
are factored in.

00:57:23.382 --> 00:57:25.173
Not, perhaps, at the
levels they should be,

00:57:25.173 --> 00:57:27.540
but some are factored in,
whereas in other countries,

00:57:27.540 --> 00:57:29.460
they're not at all.

00:57:29.460 --> 00:57:34.360
So we have what is
referred to in some terms

00:57:34.360 --> 00:57:37.400
as an uneven playing field in
the sense that in some places

00:57:37.400 --> 00:57:39.070
there are regulations
in place that

00:57:39.070 --> 00:57:40.877
increase the price
of electricity,

00:57:40.877 --> 00:57:42.710
whereas they aren't
present in other places.

00:57:42.710 --> 00:57:45.334
And in some cases, they work to
your advantage, like in Norway,

00:57:45.334 --> 00:57:47.080
for instance, or Switzerland.

00:57:47.080 --> 00:57:50.110
Big hydro countries.

00:57:50.110 --> 00:57:52.110
What are some other
potential disadvantages?

00:57:52.110 --> 00:57:56.360
You hear a lot about jobs,
jobs moving overseas.

00:57:56.360 --> 00:57:59.330
Let's start looking
at some of the numbers

00:57:59.330 --> 00:58:02.580
next Thursday, when we
have Doug coming in here

00:58:02.580 --> 00:58:05.450
and talking about
the cost model.

00:58:05.450 --> 00:58:08.160
A good number to keep
in mind, a good number

00:58:08.160 --> 00:58:11.170
to keep in mind for the
manufacturing of modules--

00:58:11.170 --> 00:58:16.890
so back to this part right
over here-- for the module

00:58:16.890 --> 00:58:21.354
itself, typical numbers would
be 2 to 12 people, depending

00:58:21.354 --> 00:58:22.770
on what region of
the world you're

00:58:22.770 --> 00:58:26.470
in, 2 to 12 people per megawatt.

00:58:26.470 --> 00:58:29.790
So if you have a total market
of, say, 10 gigawatts--

00:58:29.790 --> 00:58:30.970
it's larger than that.

00:58:30.970 --> 00:58:32.920
We'll do quick math.

00:58:32.920 --> 00:58:37.450
So if it's-- the megawatt would
be 20,000 people per gigawatt,

00:58:37.450 --> 00:58:40.580
about 10 gigawatts, somewhere
in that range, on the low end.

00:58:40.580 --> 00:58:44.050
And on the high end it would be
somewhere around 200,000 people

00:58:44.050 --> 00:58:46.590
per gigawatt, on the high
end of the labor intensity.

00:58:46.590 --> 00:58:48.173
So the high end of
the labor intensity

00:58:48.173 --> 00:58:50.340
represents a company
like Trina Solar in China

00:58:50.340 --> 00:58:52.060
that is extremely
labor intensive.

00:58:52.060 --> 00:58:53.850
It's located in Changzhou.

00:58:53.850 --> 00:58:59.606
It's on the high-speed rail
between Shanghai and Nanjing.

00:58:59.606 --> 00:59:02.230
The low end of that scale would
be more representative of, say,

00:59:02.230 --> 00:59:05.635
Suntech, which is almost next
door in Wuxi, which has adopted

00:59:05.635 --> 00:59:07.760
a different approach,
saying, we're going to invest

00:59:07.760 --> 00:59:09.754
in robots because we're
uncertain about where

00:59:09.754 --> 00:59:11.670
the whole labor thing
is going to go in China,

00:59:11.670 --> 00:59:13.045
where the prices
are going to go.

00:59:13.045 --> 00:59:15.951
Or REC, which is a
Norwegian company,

00:59:15.951 --> 00:59:18.450
but they opened a factory in
Singapore, which has very, very

00:59:18.450 --> 00:59:22.135
low labor rates, because
they invested a lot of money

00:59:22.135 --> 00:59:24.130
in capital equipment, in
robots to move modules

00:59:24.130 --> 00:59:26.020
around their factories.

00:59:26.020 --> 00:59:29.210
US, if you can look at
turnkey manufacturing lines,

00:59:29.210 --> 00:59:31.897
are typically round 3,
3.5 people per megawatt.

00:59:31.897 --> 00:59:33.730
Those are some good
numbers to keep in mind.

00:59:33.730 --> 00:59:37.240
So when anybody comes to you
and says, jobs, solar jobs,

00:59:37.240 --> 00:59:40.210
with these sorts of numbers
you can begin estimating, OK,

00:59:40.210 --> 00:59:41.320
give me a number.

00:59:41.320 --> 00:59:43.660
How big is the solar
market going to grow to?

00:59:43.660 --> 00:59:45.190
What is the level of automation?

00:59:45.190 --> 00:59:48.540
What is the labor intensity of
manufacturing of the modules?

00:59:48.540 --> 00:59:50.540
And you can make an
estimate on the total number

00:59:50.540 --> 00:59:53.120
of jobs, direct jobs,
direct jobs associated

00:59:53.120 --> 00:59:54.440
with the module manufacturing.

00:59:54.440 --> 00:59:56.170
Then there are all the
indirect jobs as well.

00:59:56.170 --> 00:59:58.169
There are all the people
supplying the commodity

00:59:58.169 --> 01:00:00.290
materials into your
manufacturing line.

01:00:00.290 --> 01:00:02.270
There all the people
who are working

01:00:02.270 --> 01:00:05.730
as administrative assistants
and R&D staff and so forth.

01:00:05.730 --> 01:00:07.870
So it's not quite
as simple as that,

01:00:07.870 --> 01:00:10.852
but it gives you a number
to really start grabbing.

01:00:10.852 --> 01:00:13.310
And that, of course, doesn't
include all the installations,

01:00:13.310 --> 01:00:15.400
installer jobs.

01:00:15.400 --> 01:00:15.900
All right.

01:00:15.900 --> 01:00:19.850
So we're talking about
grid-tied electricity.

01:00:19.850 --> 01:00:23.270
I'd like to move on to
our next few topics here.

01:00:23.270 --> 01:00:26.220
We're talking about
grid-tied electricity.

01:00:26.220 --> 01:00:27.550
This is-- ooh, it's in German.

01:00:27.550 --> 01:00:28.830
Sorry about that.

01:00:28.830 --> 01:00:33.560
This is over a period-- "tag der
woche" means "day of the week."

01:00:33.560 --> 01:00:38.080
And this is basically
the peak power

01:00:38.080 --> 01:00:41.060
in gigawatts, instantaneous
at any given point.

01:00:41.060 --> 01:00:43.350
So we have morning,
middle of the day,

01:00:43.350 --> 01:00:44.980
night, middle of the day, night.

01:00:44.980 --> 01:00:48.010
There's a little bump right
here during prime time TV.

01:00:48.010 --> 01:00:53.020
And we have the output of PV
going from 5 gigawatt peak

01:00:53.020 --> 01:00:55.100
to 30 gigawatt
peak, you can see,

01:00:55.100 --> 01:00:58.500
beginning to eat into
the profits, essentially

01:00:58.500 --> 01:01:01.360
the peak hours.

01:01:01.360 --> 01:01:03.700
So let me differentiate
between base load and peak.

01:01:03.700 --> 01:01:06.020
Base load power
would be voom, right?

01:01:06.020 --> 01:01:08.550
Something providing a
constant power output

01:01:08.550 --> 01:01:11.050
as a function of time, something
like a nuclear power plant,

01:01:11.050 --> 01:01:12.240
coal fired power plant.

01:01:12.240 --> 01:01:14.990
Peakers would be natural
gas fired power plants that

01:01:14.990 --> 01:01:15.750
receive the call.

01:01:15.750 --> 01:01:16.990
It's like, Jessica,
you're in charge

01:01:16.990 --> 01:01:18.089
of the peaker over there.

01:01:18.089 --> 01:01:19.630
Based on my weather
report today, I'm

01:01:19.630 --> 01:01:21.579
going to need so many
gigawatts tomorrow

01:01:21.579 --> 01:01:23.120
between the hours
of 10:00 and 12:00.

01:01:23.120 --> 01:01:24.250
You think you can
turn yourself on then?

01:01:24.250 --> 01:01:25.160
I'll pay this amount.

01:01:25.160 --> 01:01:25.750
Yep.

01:01:25.750 --> 01:01:26.833
Jessica's going to battle.

01:01:26.833 --> 01:01:27.820
Yes, I can do that.

01:01:27.820 --> 01:01:30.900
PV, on the other hand, is coming
on if the sun's coming on.

01:01:30.900 --> 01:01:34.410
And to date, there isn't a
great degree of predictability.

01:01:34.410 --> 01:01:36.070
We can see if that's
going to change,

01:01:36.070 --> 01:01:37.740
based on one of our
class projects here.

01:01:37.740 --> 01:01:39.823
But there wasn't a great
degree of predictability.

01:01:39.823 --> 01:01:42.140
You can see here,
for example, and here

01:01:42.140 --> 01:01:45.040
varying amounts of sunlight
from one day to the other

01:01:45.040 --> 01:01:48.230
as a result of
changing insulation.

01:01:48.230 --> 01:01:50.720
There's, of course, changing
demand as a day of the week

01:01:50.720 --> 01:01:51.220
goes by.

01:01:51.220 --> 01:01:54.470
On Saturday and Sunday there's
less demand for electricity.

01:01:54.470 --> 01:01:56.570
So what PV is doing
is eating into some

01:01:56.570 --> 01:01:58.790
of the highest-use
periods, which

01:01:58.790 --> 01:02:02.690
is good from the market's
perspective-- from the, shall

01:02:02.690 --> 01:02:05.120
we say, the "person in
charge of maintaining grid

01:02:05.120 --> 01:02:08.517
stabilities" perspective,
because they want power

01:02:08.517 --> 01:02:09.350
to be produced then.

01:02:09.350 --> 01:02:11.433
They don't want to run out
of generation capacity.

01:02:11.433 --> 01:02:14.110
And PV is avoiding
the situation,

01:02:14.110 --> 01:02:16.550
or at least prolonging the
situation a few more years,

01:02:16.550 --> 01:02:18.700
until we run out
of power generation

01:02:18.700 --> 01:02:21.800
capacity on the grid.

01:02:21.800 --> 01:02:24.530
As our population grows and
our energy intensity grows,

01:02:24.530 --> 01:02:25.960
we're needing more
and more power,

01:02:25.960 --> 01:02:27.584
but we're not building
new power plants

01:02:27.584 --> 01:02:31.050
at that rate, generally, in
Germany and the United States.

01:02:31.050 --> 01:02:33.540
And so PV is helping
to defray some

01:02:33.540 --> 01:02:36.180
of that deferred investment
in new generation capacity.

01:02:36.180 --> 01:02:37.140
So that's good.

01:02:37.140 --> 01:02:41.310
On the bad side, somebody
who has invested--

01:02:41.310 --> 01:02:45.430
let's say Mary's invested in a
natural gas generation plant.

01:02:45.430 --> 01:02:47.280
She put the money down
for Jessica's plant.

01:02:47.280 --> 01:02:48.420
Jessica's running it.

01:02:48.420 --> 01:02:51.800
And now Mary realizes,
well goodness,

01:02:51.800 --> 01:02:54.110
I thought I was going to
get a steady rate of return

01:02:54.110 --> 01:02:56.760
over 20 years from my natural
gas fired power plant.

01:02:56.760 --> 01:02:59.559
That rate of return was factored
into my economic analysis

01:02:59.559 --> 01:03:01.850
when they borrowed money at
a certain rate from a bank,

01:03:01.850 --> 01:03:04.440
and now you're telling me
that this upstart technology,

01:03:04.440 --> 01:03:06.670
this new PV stuff,
is coming online

01:03:06.670 --> 01:03:09.320
and is going to take
away some of my profit.

01:03:09.320 --> 01:03:11.510
That doesn't work for me.

01:03:11.510 --> 01:03:15.680
As a matter of fact, on
peak days in Bavaria,

01:03:15.680 --> 01:03:18.220
which is the southeastern
part of Germany,

01:03:18.220 --> 01:03:21.970
today PV can comprise as
much as 40% of electricity

01:03:21.970 --> 01:03:25.470
on the grid during peak
hours during the summer.

01:03:25.470 --> 01:03:31.220
And what that does sometimes is
force conventional electricity

01:03:31.220 --> 01:03:34.820
producers to either turn off
or go into negative pricing,

01:03:34.820 --> 01:03:36.429
meaning they shunt
the power to ground

01:03:36.429 --> 01:03:37.970
so that they don't
have to pay to put

01:03:37.970 --> 01:03:39.880
their electricity on the grid.

01:03:39.880 --> 01:03:42.460
So that's an interesting
market dynamic

01:03:42.460 --> 01:03:45.340
that's evolving as more and
more PV electricity enters

01:03:45.340 --> 01:03:49.430
the grid in these regions of
high penetration, which include

01:03:49.430 --> 01:03:52.520
California, Hawaii-- the
island of Lanai, I believe,

01:03:52.520 --> 01:03:56.990
has up to 40% as
well on weekdays.

01:03:56.990 --> 01:04:01.270
Bavaria in Germany, which is the
southeastern part of Germany.

01:04:01.270 --> 01:04:04.130
Let me get to this
point right here.

01:04:04.130 --> 01:04:05.870
Predicting where
the market will go,

01:04:05.870 --> 01:04:07.560
we talked about the
demand curve for PV,

01:04:07.560 --> 01:04:12.524
and we collapsed both
insulation and price

01:04:12.524 --> 01:04:14.440
of electricity from the
grid into this dollars

01:04:14.440 --> 01:04:16.180
per watt-peak figure, which
is how much people are

01:04:16.180 --> 01:04:17.130
willing to pay.

01:04:17.130 --> 01:04:20.520
We can also break it out into
annual solar energy yield.

01:04:20.520 --> 01:04:22.980
This is the solar
resource available.

01:04:22.980 --> 01:04:26.250
It's using the units of kilowatt
hours per kilowatt peak,

01:04:26.250 --> 01:04:27.910
meaning the number
of kilowatt hours,

01:04:27.910 --> 01:04:30.260
the amount of
energy, that a unit

01:04:30.260 --> 01:04:32.857
of rated power, kilowatt
peak, of the solar panel

01:04:32.857 --> 01:04:33.690
is going to produce.

01:04:33.690 --> 01:04:36.800
So how much energy will a
unit of solar panel produce?

01:04:36.800 --> 01:04:37.720
That's down here.

01:04:37.720 --> 01:04:39.470
And of course,
larger amounts mean

01:04:39.470 --> 01:04:41.720
more sun, more solar resource.

01:04:41.720 --> 01:04:45.500
On the ordinate, we have average
power price of household.

01:04:45.500 --> 01:04:47.440
This is US dollars.

01:04:47.440 --> 01:04:50.400
It should be average
electricity price, US dollars

01:04:50.400 --> 01:04:51.315
per kilowatt hour.

01:04:51.315 --> 01:04:53.690
So that's how much people are
paying for the electricity,

01:04:53.690 --> 01:04:54.897
what you're displacing.

01:04:54.897 --> 01:04:57.230
And you can see that these
little bubbles here represent

01:04:57.230 --> 01:04:58.188
the size of the market.

01:04:58.188 --> 01:05:02.490
So as the price of PV
electricity comes down,

01:05:02.490 --> 01:05:05.270
we are able to address
more and more markets.

01:05:05.270 --> 01:05:09.590
As the subsidization of
PV electricity increases,

01:05:09.590 --> 01:05:11.580
you move these little bubble up.

01:05:11.580 --> 01:05:13.410
And as the subsidization
comes back down,

01:05:13.410 --> 01:05:15.610
the market pull-- not the
manufacturing subsidies

01:05:15.610 --> 01:05:19.400
but the use or
installation subsidies--

01:05:19.400 --> 01:05:22.660
as they come back down, the
little bubbles move down,

01:05:22.660 --> 01:05:25.800
and eventually rest at
their so-called true market

01:05:25.800 --> 01:05:27.697
price without the
externalities factored in.

01:05:27.697 --> 01:05:29.280
And so you can see
how we're beginning

01:05:29.280 --> 01:05:32.860
to enter a regime where PV is
starting to look interesting

01:05:32.860 --> 01:05:33.962
in a lot of places.

01:05:33.962 --> 01:05:35.670
And as we begin
addressing these markets,

01:05:35.670 --> 01:05:38.270
or hitting these markets, you
can imagine two Gaussian curves

01:05:38.270 --> 01:05:40.664
intersecting, one being the
price of PV electricity,

01:05:40.664 --> 01:05:42.580
the other being the price
of grid electricity,

01:05:42.580 --> 01:05:44.176
as they intersect and overlap.

01:05:44.176 --> 01:05:45.550
You expect an
exponential growth,

01:05:45.550 --> 01:05:47.210
and that's what we're
seeing in the market today.

01:05:47.210 --> 01:05:47.630
Question?

01:05:47.630 --> 01:05:49.463
AUDIENCE: What is the
California [INAUDIBLE]

01:05:49.463 --> 01:05:50.384
for tier 4 and tier 5?

01:05:50.384 --> 01:05:51.050
PROFESSOR: Yeah.

01:05:51.050 --> 01:05:56.400
So California decided
to implement more of a--

01:05:56.400 --> 01:06:00.380
well, it is a little bit
regressive in that sense.

01:06:00.380 --> 01:06:03.840
But what they did is,
they said let's, instead

01:06:03.840 --> 01:06:06.295
of paying the same
price for electricity

01:06:06.295 --> 01:06:09.245
for everybody--
which actually it

01:06:09.245 --> 01:06:10.880
would be even the opposite way.

01:06:10.880 --> 01:06:13.300
You know how residential, we
typically pay higher rates

01:06:13.300 --> 01:06:14.340
than industrial?

01:06:14.340 --> 01:06:16.020
That's because
industrial buys in bulk.

01:06:16.020 --> 01:06:18.010
They buy electricity
Costco size.

01:06:18.010 --> 01:06:20.799
We just buy it
corner-store size.

01:06:20.799 --> 01:06:22.590
And so they pay less
for their electricity.

01:06:22.590 --> 01:06:24.870
And California realized
this and said, well, this

01:06:24.870 --> 01:06:26.419
is a perverse incentive.

01:06:26.419 --> 01:06:28.460
What this does is it gives
an incentive to people

01:06:28.460 --> 01:06:29.834
to use more electricity.

01:06:29.834 --> 01:06:31.250
Because if you buy
in Costco size,

01:06:31.250 --> 01:06:32.720
you pay less per
kilowatt hour than

01:06:32.720 --> 01:06:34.430
if you buy in the corner store.

01:06:34.430 --> 01:06:37.830
Instead, what we're going to
do is reverse that incentive

01:06:37.830 --> 01:06:41.250
by penalizing the people
who buy more electricity,

01:06:41.250 --> 01:06:43.650
so your rate is going to
be higher then if you just

01:06:43.650 --> 01:06:45.405
bought a small amount.

01:06:45.405 --> 01:06:47.530
So what they did is, the
government bureaucrats got

01:06:47.530 --> 01:06:49.640
together with some of
the scholars and decided,

01:06:49.640 --> 01:06:54.590
these are reasonable amounts for
a residential house to consume.

01:06:54.590 --> 01:06:56.910
This is the next tier up,
this is the next tier up,

01:06:56.910 --> 01:06:59.900
this is next tier up, and this
is completely unreasonable.

01:06:59.900 --> 01:07:03.180
And so what they did is they
tiered their electricity.

01:07:03.180 --> 01:07:06.780
And you have California
tiers 1, 2, and 3 represented

01:07:06.780 --> 01:07:10.190
by the big bubble, and
tier 4 and tier 5 as

01:07:10.190 --> 01:07:11.630
represented at higher pricing.

01:07:11.630 --> 01:07:14.210
So if you happen to fall
into the tier 5 category,

01:07:14.210 --> 01:07:17.050
you could be paying upward
of $0.30 per kilowatt hour

01:07:17.050 --> 01:07:19.150
for your electricity.

01:07:19.150 --> 01:07:22.830
And it's a way of penalizing
you for powering two swimming

01:07:22.830 --> 01:07:26.400
pools in your backyard
along with, I don't know,

01:07:26.400 --> 01:07:28.820
something in the range of
10 kilowatts of lighting

01:07:28.820 --> 01:07:30.630
inside of your house.

01:07:30.630 --> 01:07:33.830
Whereas, if you're running
on a meager budget,

01:07:33.830 --> 01:07:36.130
you might do much better.

01:07:36.130 --> 01:07:37.522
Do you have a
question over here?

01:07:37.522 --> 01:07:39.715
AUDIENCE: Is it
only residential?

01:07:39.715 --> 01:07:42.040
PROFESSOR: California's
tier 1, 2, 3, and 4, and 5?

01:07:42.040 --> 01:07:43.380
That a great question.

01:07:43.380 --> 01:07:46.070
Joe, would you happen to know
if commercial and industrial

01:07:46.070 --> 01:07:48.853
fall into tiers as well, or
if it's just residential?

01:07:48.853 --> 01:07:50.340
AUDIENCE: My guess would be no,
because that would be political

01:07:50.340 --> 01:07:50.770
suicide.

01:07:50.770 --> 01:07:51.955
PROFESSOR: Well,
even the residential

01:07:51.955 --> 01:07:52.975
is political suicide.

01:07:52.975 --> 01:07:53.600
AUDIENCE: Yeah.

01:07:53.600 --> 01:07:55.899
The residential system is
actually kind of a bad one,

01:07:55.899 --> 01:07:57.440
because if you have
like eight people

01:07:57.440 --> 01:08:00.106
living in a house versus two, it
doesn't take that into account.

01:08:00.106 --> 01:08:04.650
So it's not a per capita
thing. [INAUDIBLE].

01:08:04.650 --> 01:08:07.510
PROFESSOR: So many
states and countries

01:08:07.510 --> 01:08:10.780
have tried to adjust
incentive mechanisms.

01:08:10.780 --> 01:08:12.300
And I'm not sure
specifically about

01:08:12.300 --> 01:08:13.712
the commercial and industrial.

01:08:13.712 --> 01:08:15.920
You might want to jot it
down and give a quick Google

01:08:15.920 --> 01:08:18.729
search after.

01:08:18.729 --> 01:08:19.260
Germany.

01:08:19.260 --> 01:08:21.500
Growing gangbusters.

01:08:21.500 --> 01:08:22.470
The market versus time.

01:08:22.470 --> 01:08:25.750
This is installed capacity
annually versus year.

01:08:25.750 --> 01:08:28.880
And again, subsidies, support
mechanisms, tax breaks,

01:08:28.880 --> 01:08:30.060
incentives.

01:08:30.060 --> 01:08:31.979
What we wanted to
emphasize here was

01:08:31.979 --> 01:08:34.859
that Germany is not one of the
sunnier places in the world.

01:08:34.859 --> 01:08:35.910
Look at this.

01:08:35.910 --> 01:08:38.380
It's almost half of
the modulus installed,

01:08:38.380 --> 01:08:44.479
and yet it has less insulation
than even here in the US

01:08:44.479 --> 01:08:45.010
northeast.

01:08:45.010 --> 01:08:46.630
Even the sunniest
places in Germany.

01:08:46.630 --> 01:08:47.529
This is Bavaria.

01:08:47.529 --> 01:08:50.880
This is the place that
hits 40% of electricity

01:08:50.880 --> 01:08:53.170
on peak summer days
coming from PV.

01:08:53.170 --> 01:08:56.090
And this is where we
are, right up there.

01:08:56.090 --> 01:08:58.819
Same scale.

01:08:58.819 --> 01:09:04.529
So something's going on in
Germany, and it's not sun.

01:09:04.529 --> 01:09:07.880
The German government instituted
a number of-- this little D

01:09:07.880 --> 01:09:09.760
represents
Deutschland, Germany--

01:09:09.760 --> 01:09:13.120
implemented a number
of incentive programs

01:09:13.120 --> 01:09:17.000
to increase the amount of
PV installed on the grid,

01:09:17.000 --> 01:09:20.350
under the assumption that
if they had demand pull,

01:09:20.350 --> 01:09:21.600
they would have a supply push.

01:09:21.600 --> 01:09:24.040
Meaning, they would start
up local industries.

01:09:24.040 --> 01:09:26.580
Initially it didn't quite take
off as fast as they hoped,

01:09:26.580 --> 01:09:29.540
so they began providing
direct manufacturing support,

01:09:29.540 --> 01:09:33.590
direct subsidies to install
PV manufacturing plants,

01:09:33.590 --> 01:09:36.189
to the point where they
were willing to pay $0.50

01:09:36.189 --> 01:09:39.529
on the euro for each piece of
capital equipment that moved

01:09:39.529 --> 01:09:42.950
in on German soil to produce
plants in the mid 2000s.

01:09:42.950 --> 01:09:45.399
So that combination
of events left them

01:09:45.399 --> 01:09:48.649
with a booming market,
upwards of 100,000 jobs,

01:09:48.649 --> 01:09:51.439
I believe, in Germany
related to PV manufacturing,

01:09:51.439 --> 01:09:54.380
and a lot of PV
installed on their grid

01:09:54.380 --> 01:09:56.240
as a result of the market pull.

01:09:56.240 --> 01:09:58.700
So really, when it comes
to subsidies and support,

01:09:58.700 --> 01:10:01.470
you have to decouple what
is market pull, which

01:10:01.470 --> 01:10:03.390
means we want PV
installed on our grid

01:10:03.390 --> 01:10:07.210
to offset our country's
carbon emissions,

01:10:07.210 --> 01:10:09.429
versus manufacturing
push, which is to say,

01:10:09.429 --> 01:10:11.220
you want to set up a
factory in my country?

01:10:11.220 --> 01:10:11.719
Fine.

01:10:11.719 --> 01:10:12.320
Come right in.

01:10:12.320 --> 01:10:13.590
I'll provide you the finances.

01:10:13.590 --> 01:10:15.840
I'll provide you a lower
cost of capital on your loan,

01:10:15.840 --> 01:10:20.760
or help you by providing
a direct grant or loan.

01:10:20.760 --> 01:10:24.170
And these are two different
support mechanisms,

01:10:24.170 --> 01:10:26.960
and that's why you see that
dichotomy between Germany,

01:10:26.960 --> 01:10:28.730
which installs a
lot of PV and has

01:10:28.730 --> 01:10:30.270
a decent amount
of manufacturing;

01:10:30.270 --> 01:10:32.350
the US, which has a
lot of installation

01:10:32.350 --> 01:10:34.890
but not a whole lot of
manufacturing traditionally,

01:10:34.890 --> 01:10:36.990
although it's changed,
the support has changed

01:10:36.990 --> 01:10:39.111
over the last three
years; and China, which

01:10:39.111 --> 01:10:40.610
has a heck of a lot
of manufacturing

01:10:40.610 --> 01:10:42.540
but not a whole lot
of local demand.

01:10:42.540 --> 01:10:45.350
It's because each country has
adopted a different approach

01:10:45.350 --> 01:10:48.270
based on optimization
of different functions.

01:10:48.270 --> 01:10:52.600
In terms of where a lot has
happened on the state level,

01:10:52.600 --> 01:10:54.940
these are RPS, renewable
portfolio standard,

01:10:54.940 --> 01:10:57.790
policies with solar and
distributed generation

01:10:57.790 --> 01:11:00.530
provisions distributed
throughout the United States.

01:11:00.530 --> 01:11:02.660
And you can see that the
states have picked up

01:11:02.660 --> 01:11:06.060
where the federal government
lacked in terms of leadership.

01:11:06.060 --> 01:11:07.560
In terms of getting
PV on the grid,

01:11:07.560 --> 01:11:10.020
the states really pulled hard.

01:11:10.020 --> 01:11:13.700
In, specifically, New Jersey--
some really oddball states,

01:11:13.700 --> 01:11:17.490
like New Jersey here, installing
about 5 gigawatts of solar.

01:11:17.490 --> 01:11:23.910
California had another more
of a market pull mechanism.

01:11:23.910 --> 01:11:25.470
Rebate programs for
renewables, this

01:11:25.470 --> 01:11:28.840
is if you install solar panels
you might get a few thousand

01:11:28.840 --> 01:11:31.780
dollars back from
the state government.

01:11:31.780 --> 01:11:35.990
And it just gives you a
sense of how the states have

01:11:35.990 --> 01:11:40.160
filled in that void where
the federal government didn't

01:11:40.160 --> 01:11:40.975
step in.

01:11:40.975 --> 01:11:42.850
And this has come, in
part, to our detriment.

01:11:42.850 --> 01:11:45.880
It's nice, because we can
allow-- the federalist

01:11:45.880 --> 01:11:48.450
approach allows individual
states to say, hey,

01:11:48.450 --> 01:11:49.541
we prioritize green jobs.

01:11:49.541 --> 01:11:50.040
All right.

01:11:50.040 --> 01:11:51.000
Let's support it.

01:11:51.000 --> 01:11:52.090
Let's make it happen.

01:11:52.090 --> 01:11:53.840
And that's certainly
Massachusetts' take.

01:11:53.840 --> 01:12:00.030
California, Texas, certain
other states throughout the US,

01:12:00.030 --> 01:12:02.660
including Mississippi
most recently.

01:12:02.660 --> 01:12:05.080
But the downside
of that is that you

01:12:05.080 --> 01:12:07.850
have each state doing its
own thing, each municipality,

01:12:07.850 --> 01:12:11.750
each local electric utility
grid doing its own thing.

01:12:11.750 --> 01:12:15.260
And what can happen
is diversification

01:12:15.260 --> 01:12:17.760
of support mechanisms and
incentives to the point

01:12:17.760 --> 01:12:19.800
now that if you want
to install panels

01:12:19.800 --> 01:12:23.820
on your roof in the United
States, there over 18,000 ways

01:12:23.820 --> 01:12:25.840
you could do it
within the US alone.

01:12:25.840 --> 01:12:29.030
18,000 unique types
of paperwork that you

01:12:29.030 --> 01:12:30.980
have to fill out,
depending on where

01:12:30.980 --> 01:12:32.290
you are in the United States.

01:12:32.290 --> 01:12:35.360
In Germany, you have
one unified paperwork

01:12:35.360 --> 01:12:38.340
for the entire country,
and it's very short.

01:12:38.340 --> 01:12:39.910
I believe it's two pages.

01:12:39.910 --> 01:12:41.780
And in the United
States, you can fill out

01:12:41.780 --> 01:12:43.390
a few reams of paper.

01:12:43.390 --> 01:12:45.330
One of our friends
in California once

01:12:45.330 --> 01:12:49.570
estimated it takes about $1
a watt to fill out paperwork

01:12:49.570 --> 01:12:51.315
and to get certifications
and so forth.

01:12:51.315 --> 01:12:53.190
Maybe it was a little
bit of an exaggeration,

01:12:53.190 --> 01:12:54.940
but the price is high.

01:12:54.940 --> 01:12:56.500
And so recently the
DOE has attempted

01:12:56.500 --> 01:12:58.540
to address this as
well by fostering

01:12:58.540 --> 01:13:00.800
a "race to the
top" type incentive

01:13:00.800 --> 01:13:04.050
mechanism for states to
facilitate installation.

01:13:04.050 --> 01:13:06.535
But still, in Massachusetts,
we had-- remember

01:13:06.535 --> 01:13:08.910
for the system on my roof, we
had at least two inspectors

01:13:08.910 --> 01:13:11.010
come by from different agencies.

01:13:11.010 --> 01:13:13.420
That could have been
consolidated into one.

01:13:13.420 --> 01:13:17.050
So while state
independence is good,

01:13:17.050 --> 01:13:18.932
and local utility
independence is great,

01:13:18.932 --> 01:13:21.390
from the perspective that the
federal government is lacking

01:13:21.390 --> 01:13:23.090
in terms of moving
forward, the states

01:13:23.090 --> 01:13:25.730
can push forward at
their comfortable speed.

01:13:25.730 --> 01:13:28.870
On the downside, it makes it
very difficult for an installer

01:13:28.870 --> 01:13:31.940
company to reduce
its costs because it

01:13:31.940 --> 01:13:34.940
has to address a
multitude of local markets

01:13:34.940 --> 01:13:36.100
throughout the US.

01:13:36.100 --> 01:13:38.690
And that adds to cost.

01:13:38.690 --> 01:13:40.850
In terms of projections,
I really hesitate.

01:13:40.850 --> 01:13:44.210
Because in 2006, there were
a number projections of where

01:13:44.210 --> 01:13:48.350
we'd be in 2010,
and many of them--

01:13:48.350 --> 01:13:51.180
you can see the range here,
going from 4 gigawatts up to,

01:13:51.180 --> 01:13:53.600
actually, I think Photon
Consulting had us close

01:13:53.600 --> 01:13:58.680
to 30 gigawatts back in 2010.

01:13:58.680 --> 01:14:00.310
Photon Consulting
was closer to right

01:14:00.310 --> 01:14:02.920
than most of the other
groups, interestingly.

01:14:02.920 --> 01:14:05.370
The 4 gigawatts was
obviously pretty far off.

01:14:05.370 --> 01:14:07.520
So projections are very
dangerous because you

01:14:07.520 --> 01:14:09.040
could be easily wrong.

01:14:09.040 --> 01:14:12.337
This is just to
show you an example.

01:14:12.337 --> 01:14:14.420
These are projections of
where the solar market is

01:14:14.420 --> 01:14:15.750
going to go.

01:14:15.750 --> 01:14:18.630
From the EPIA, which is the
European Photovoltaics Industry

01:14:18.630 --> 01:14:21.330
Association, it's one
of the better sources

01:14:21.330 --> 01:14:23.840
that I could find with the
various scenarios giving you

01:14:23.840 --> 01:14:25.830
an indication, a tornado
plot, if you will,

01:14:25.830 --> 01:14:32.170
of where the market is going
to head over the future.

01:14:32.170 --> 01:14:34.610
And the bottom line
is that most likely

01:14:34.610 --> 01:14:35.690
it will continue to grow.

01:14:35.690 --> 01:14:37.731
Even though it's going to
enter a difficult year,

01:14:37.731 --> 01:14:39.860
a difficult two years as
the oversupply condition

01:14:39.860 --> 01:14:41.600
works itself out of
the market, we're

01:14:41.600 --> 01:14:43.191
headed for some
pretty massive growth.

01:14:43.191 --> 01:14:44.690
And just keep those
numbers in mind.

01:14:44.690 --> 01:14:49.200
If this grows up to be, say,
a 100-gigawatt industry,

01:14:49.200 --> 01:14:50.910
and there's two
people per megawatt,

01:14:50.910 --> 01:14:53.670
that's 200,000
people employed just

01:14:53.670 --> 01:14:56.470
in the manufacturing of modules.

01:14:56.470 --> 01:14:58.600
Not counting all
the other people

01:14:58.600 --> 01:15:00.760
on top of that, maybe
another order of magnitude,

01:15:00.760 --> 01:15:03.420
if we use the car
industry as an example.

01:15:03.420 --> 01:15:06.830
So we could be looking at
two million jobs worldwide.

01:15:06.830 --> 01:15:11.180
And where they are, who works
on them, how the value chain is

01:15:11.180 --> 01:15:14.990
distributed between
R&D, manufacturing,

01:15:14.990 --> 01:15:17.326
this is all open.

01:15:17.326 --> 01:15:18.700
The next decade
will decide that.

01:15:18.700 --> 01:15:20.080
You'll help decide that.

01:15:20.080 --> 01:15:24.140
So with that message,
I let you go.