1
00:00:00,060 --> 00:00:02,490
The following content is
provided under a Creative

2
00:00:02,490 --> 00:00:04,010
Commons license.

3
00:00:04,010 --> 00:00:06,340
Your support will help
MIT OpenCourseWare

4
00:00:06,340 --> 00:00:10,710
continue to offer high quality
educational resources for free.

5
00:00:10,710 --> 00:00:13,320
To make a donation or
view additional materials

6
00:00:13,320 --> 00:00:17,213
from hundreds of MIT courses,
visit MIT OpenCourseWare

7
00:00:17,213 --> 00:00:17,838
at ocw.mit.edu.

8
00:00:25,760 --> 00:00:28,790
PROFESSOR: So modules,
systems, and reliability.

9
00:00:28,790 --> 00:00:32,140
What we're going to do is talk
about how we go from the cells,

10
00:00:32,140 --> 00:00:35,520
or from the films,
to full modules

11
00:00:35,520 --> 00:00:37,294
and, finally, to systems.

12
00:00:37,294 --> 00:00:38,710
So our first
learning objective is

13
00:00:38,710 --> 00:00:41,890
to describe, more or less, the
DNA or anatomy of a PV module.

14
00:00:41,890 --> 00:00:45,285
This will be a bit of a prep
for the visit to Fraunhofer

15
00:00:45,285 --> 00:00:47,630
CSE, where we'll
get to see the labs

16
00:00:47,630 --> 00:00:50,410
and see all the materials
that go into a PV module.

17
00:00:50,410 --> 00:00:52,140
So just establishing
some definitions

18
00:00:52,140 --> 00:00:54,660
up front so we're all
speaking the same language--

19
00:00:54,660 --> 00:00:56,181
you start out with
either a film.

20
00:00:56,181 --> 00:00:56,680
Right?

21
00:00:56,680 --> 00:00:58,630
If you're depositing
a thin film material,

22
00:00:58,630 --> 00:01:01,260
which is usually divided
into discrete devices

23
00:01:01,260 --> 00:01:04,489
using laser processing,
or you can start out

24
00:01:04,489 --> 00:01:07,430
with a discrete wafer which
has been processed into a cell.

25
00:01:07,430 --> 00:01:10,010
The module is the
combination of many

26
00:01:10,010 --> 00:01:12,859
of these cells, shown
very generically right

27
00:01:12,859 --> 00:01:14,150
here without the interconnects.

28
00:01:14,150 --> 00:01:16,312
Because you can tie
up the cells together

29
00:01:16,312 --> 00:01:17,770
in parallel or in
series, depending

30
00:01:17,770 --> 00:01:20,820
on what sort of voltage
and current outputs

31
00:01:20,820 --> 00:01:22,520
you desire from your module.

32
00:01:22,520 --> 00:01:25,810
And then finally the modules are
situated together in an array.

33
00:01:25,810 --> 00:01:29,730
That's showing a combination
of six modules forming

34
00:01:29,730 --> 00:01:31,560
that array right there.

35
00:01:31,560 --> 00:01:35,260
And as you can see, modules are
typically this size right here.

36
00:01:35,260 --> 00:01:39,120
They're usually about this
size or even a little larger.

37
00:01:39,120 --> 00:01:42,280
They're rather heavy,
so you can get a sense.

38
00:01:42,280 --> 00:01:44,426
They're not extremely light.

39
00:01:44,426 --> 00:01:45,884
The majority of
that mass is coming

40
00:01:45,884 --> 00:01:49,370
from the glass in the front,
some from the aluminum as well.

41
00:01:49,370 --> 00:01:51,420
And they're comprised
of many materials.

42
00:01:51,420 --> 00:01:54,190
If you were to take this
apart-- which we'll actually

43
00:01:54,190 --> 00:01:56,334
see the individual components
at Fraunhofer CSE--

44
00:01:56,334 --> 00:01:58,500
you'll notice that there
are many materials involved

45
00:01:58,500 --> 00:02:00,162
in making one of these modules.

46
00:02:00,162 --> 00:02:02,120
From the back skin
materials, the junction box,

47
00:02:02,120 --> 00:02:05,480
the aluminum framing on the
side, the glass on the front,

48
00:02:05,480 --> 00:02:07,910
the encapsulant materials that
fuse everything together--

49
00:02:07,910 --> 00:02:10,729
that bind everything--
and the cells,

50
00:02:10,729 --> 00:02:12,970
the solar cell
devices themselves.

51
00:02:12,970 --> 00:02:17,419
So our first step is to
walk through the anatomy

52
00:02:17,419 --> 00:02:20,085
of a module, at least in theory,
so that when we actually see it

53
00:02:20,085 --> 00:02:22,680
at Fraunhofer CSE, we'll
have a better sense of what

54
00:02:22,680 --> 00:02:23,790
we're looking at that.

55
00:02:23,790 --> 00:02:23,910
Yeah.

56
00:02:23,910 --> 00:02:24,410
Ashley?

57
00:02:24,410 --> 00:02:26,787
AUDIENCE: Here, what
determines that specific size?

58
00:02:26,787 --> 00:02:28,620
We've already talked
about the pseudo-square

59
00:02:28,620 --> 00:02:32,811
and how you have to balance
those two different cost

60
00:02:32,811 --> 00:02:33,310
parameters.

61
00:02:33,310 --> 00:02:35,917
But why that size specifically?

62
00:02:35,917 --> 00:02:37,500
PROFESSOR: Why that
size specifically?

63
00:02:37,500 --> 00:02:39,920
We're going to get to
that in a few slides.

64
00:02:39,920 --> 00:02:41,490
It comes, as you
probably guessed,

65
00:02:41,490 --> 00:02:43,197
from some historical reasons.

66
00:02:43,197 --> 00:02:44,420
AUDIENCE: OK.

67
00:02:44,420 --> 00:02:46,980
PROFESSOR: So some
basic principles

68
00:02:46,980 --> 00:02:49,900
about solar modules-- there's
quite a diversity of modules

69
00:02:49,900 --> 00:02:51,390
when you look at them.

70
00:02:51,390 --> 00:02:54,370
These are examples just
within one company,

71
00:02:54,370 --> 00:02:57,930
just to highlight the
range of module types

72
00:02:57,930 --> 00:02:59,860
that you might see.

73
00:02:59,860 --> 00:03:02,830
Right here, you can see
in this particular module,

74
00:03:02,830 --> 00:03:04,850
there's very little
anti-reflective coating

75
00:03:04,850 --> 00:03:05,760
on the glass.

76
00:03:05,760 --> 00:03:08,440
So you're able to look
through the front glass

77
00:03:08,440 --> 00:03:11,870
and see some white spaces
in between the cells.

78
00:03:11,870 --> 00:03:14,880
We know, from the lecture
on crystalline silicon PV,

79
00:03:14,880 --> 00:03:17,580
that when you see these little
gaps in between the cells,

80
00:03:17,580 --> 00:03:18,720
those are pseudo-squares.

81
00:03:18,720 --> 00:03:19,220
Right?

82
00:03:19,220 --> 00:03:23,060
Those are those round ingots
of Czochralski silicon

83
00:03:23,060 --> 00:03:24,660
that have been
sliced into wafers.

84
00:03:24,660 --> 00:03:27,000
And then the edges
have been shaven off,

85
00:03:27,000 --> 00:03:29,460
and you have that stop sign
type pattern of a solar cell.

86
00:03:29,460 --> 00:03:30,440
Right?

87
00:03:30,440 --> 00:03:33,820
So that way you know, OK, these
are single crystalline cells

88
00:03:33,820 --> 00:03:35,560
lined up in a module together.

89
00:03:35,560 --> 00:03:37,590
And there's very little
anti-reflective coating

90
00:03:37,590 --> 00:03:39,131
on the glass, which
means that you're

91
00:03:39,131 --> 00:03:41,026
losing a fair amount
due to reflectance.

92
00:03:41,026 --> 00:03:42,650
The module on the
right-hand side right

93
00:03:42,650 --> 00:03:47,140
here has that anti-reflection
coating on the glass.

94
00:03:47,140 --> 00:03:50,910
It could be an index
of refraction gradient

95
00:03:50,910 --> 00:03:52,240
to absorb more of the light.

96
00:03:52,240 --> 00:03:55,270
It could also be surface
texturing to scatter the light

97
00:03:55,270 --> 00:03:58,660
and trap it by total internal
reflection once it gets inside.

98
00:03:58,660 --> 00:04:00,980
There's a few different
ways that one can do that.

99
00:04:00,980 --> 00:04:04,890
Usually with a film
coating is the most common.

100
00:04:04,890 --> 00:04:06,720
And here on the
left-hand side, we

101
00:04:06,720 --> 00:04:10,190
have a similar
sub-cell component

102
00:04:10,190 --> 00:04:13,560
but much larger module comprised
of a larger number of cells

103
00:04:13,560 --> 00:04:16,300
with a higher power output.

104
00:04:16,300 --> 00:04:18,910
And we can see quite
a wide variance

105
00:04:18,910 --> 00:04:20,240
of different types of modules.

106
00:04:20,240 --> 00:04:24,380
That will be an important
message in a few.

107
00:04:24,380 --> 00:04:29,207
The water rinse cycle really
depends on where you're at.

108
00:04:29,207 --> 00:04:31,290
For the modules on my roof,
we just leave them up.

109
00:04:31,290 --> 00:04:33,850
Here in New England,
it's pretty much OK.

110
00:04:33,850 --> 00:04:36,830
If we did have trees that
deposited leaves nearby,

111
00:04:36,830 --> 00:04:40,140
we'd have to clean it off
in the fall, but we don't.

112
00:04:40,140 --> 00:04:41,850
If you're mounting
modules in the Middle

113
00:04:41,850 --> 00:04:44,920
East, where that
fine-grained sand exists

114
00:04:44,920 --> 00:04:47,960
and, if you're within 10
kilometers from the coast

115
00:04:47,960 --> 00:04:50,620
where you have the salt and
the seawater in the air,

116
00:04:50,620 --> 00:04:54,240
you can get this really
hard-to-remove sand

117
00:04:54,240 --> 00:04:55,330
caking the modules.

118
00:04:55,330 --> 00:04:59,490
And you would have to clean them
much more frequently than this.

119
00:04:59,490 --> 00:05:01,640
As a matter of fact, I
believe the numbers are--

120
00:05:01,640 --> 00:05:09,400
in Abu Dhabi, the capital of
one of the Emirates in the UAE,

121
00:05:09,400 --> 00:05:14,690
if you were to just leave your
modules out near the ocean,

122
00:05:14,690 --> 00:05:18,260
near the desert, you
would have a 40% drop

123
00:05:18,260 --> 00:05:21,230
in module output over a month.

124
00:05:21,230 --> 00:05:24,090
So that's what happens
in certain environments.

125
00:05:24,090 --> 00:05:26,050
In certain others,
like New England,

126
00:05:26,050 --> 00:05:29,530
we're more buffered
in that regard.

127
00:05:29,530 --> 00:05:32,330
We have frequent rain,
and we don't have the dust

128
00:05:32,330 --> 00:05:34,530
coming from any nearby desert.

129
00:05:34,530 --> 00:05:36,420
Typically, there
are no moving parts,

130
00:05:36,420 --> 00:05:40,040
and typically there's a 20- to
30-year manufacturer warranty.

131
00:05:40,040 --> 00:05:42,930
Some of the newer materials
that have been less tested

132
00:05:42,930 --> 00:05:45,200
might give, say, a 10-year
manufacturer's warranty

133
00:05:45,200 --> 00:05:48,620
and have to offset the
risk in years 10 to 20

134
00:05:48,620 --> 00:05:54,640
by lowering the cost up front
or lowering the price up front.

135
00:05:54,640 --> 00:05:57,300
So we have some basics
about solar modules

136
00:05:57,300 --> 00:05:58,500
just to situate ourselves.

137
00:05:58,500 --> 00:06:00,250
Let's dive into the
module DNA, since this

138
00:06:00,250 --> 00:06:03,710
is where the rubber
hits the road.

139
00:06:03,710 --> 00:06:05,690
I'm going to show
this to you in theory,

140
00:06:05,690 --> 00:06:08,450
and then we're
going to see reams

141
00:06:08,450 --> 00:06:12,020
of these module innard materials
when we visit Fraunhofer.

142
00:06:12,020 --> 00:06:15,280
So we start with the
solar cells themselves.

143
00:06:15,280 --> 00:06:18,570
These are these little
blue objects right there.

144
00:06:18,570 --> 00:06:20,070
Typically, these
cells are already

145
00:06:20,070 --> 00:06:21,670
strung together at that point.

146
00:06:21,670 --> 00:06:25,137
They have contact metallization
on the front side, which

147
00:06:25,137 --> 00:06:27,720
we deposit, say, for example,
by screen printing, which you've

148
00:06:27,720 --> 00:06:29,520
done in the lab downstairs.

149
00:06:29,520 --> 00:06:30,970
And then you string
them together

150
00:06:30,970 --> 00:06:33,340
using a machine called
a tabber-stringer

151
00:06:33,340 --> 00:06:36,240
to connect one cell to the next,
essentially the front of one

152
00:06:36,240 --> 00:06:37,880
cell to the back of the next.

153
00:06:37,880 --> 00:06:39,830
And they're all
lined up like that.

154
00:06:39,830 --> 00:06:45,530
And you deposit layers of EVA,
which is ethyl vinyl acetate.

155
00:06:45,530 --> 00:06:47,595
This is typically the
encapsulant material used.

156
00:06:47,595 --> 00:06:49,470
It's shown in red right
here in this drawing.

157
00:06:49,470 --> 00:06:53,580
It's the encapsulant material
used for crystalline silicon

158
00:06:53,580 --> 00:06:54,640
PV.

159
00:06:54,640 --> 00:06:56,900
On the back side,
there is a sheet

160
00:06:56,900 --> 00:06:58,140
of material called Tedlar.

161
00:06:58,140 --> 00:07:00,015
Despite it being drawn
in yellow right there,

162
00:07:00,015 --> 00:07:02,320
it's the white
skin material here.

163
00:07:02,320 --> 00:07:07,340
It's this white material
right around here.

164
00:07:07,340 --> 00:07:09,720
And we have the glass
on the front side.

165
00:07:09,720 --> 00:07:11,670
And typically,
it's low-iron glass

166
00:07:11,670 --> 00:07:14,640
so it can transmit
the ultraviolet light,

167
00:07:14,640 --> 00:07:17,570
or a large portion of it,
to give better blue response

168
00:07:17,570 --> 00:07:18,580
to your solar cells.

169
00:07:18,580 --> 00:07:21,761
So you're capturing a larger
portion of the solar spectrum.

170
00:07:21,761 --> 00:07:22,260
OK.

171
00:07:22,260 --> 00:07:25,630
So we have the different
components right there.

172
00:07:25,630 --> 00:07:28,760
One little piece of
trivia about EVA--

173
00:07:28,760 --> 00:07:31,470
for the chemists
in the room, what

174
00:07:31,470 --> 00:07:34,580
do you think this
would decompose into?

175
00:07:34,580 --> 00:07:39,590
Let me just give you
a little hand here.

176
00:07:39,590 --> 00:07:42,910
If you had some-- let's see.

177
00:07:47,432 --> 00:07:48,100
All right.

178
00:07:48,100 --> 00:07:51,255
So these represent
methyl groups, hydrogens,

179
00:07:51,255 --> 00:07:53,380
carbon, carbon, carbon,
double bonded to an oxygen.

180
00:07:53,380 --> 00:07:54,596
Anybody recognize?

181
00:07:54,596 --> 00:07:56,064
AUDIENCE: It looks like acetone?

182
00:07:56,064 --> 00:07:56,730
PROFESSOR: Yeah.

183
00:07:56,730 --> 00:07:59,710
So if you're looking
at the decomposition

184
00:07:59,710 --> 00:08:01,860
of this little group
right here, you

185
00:08:01,860 --> 00:08:05,140
could easily envision it
decomposing into acetic acid

186
00:08:05,140 --> 00:08:08,060
from the name ethyl
vinyl acetate.

187
00:08:08,060 --> 00:08:12,430
So if you have a thin
film material that

188
00:08:12,430 --> 00:08:14,490
could react with
the encapsulant,

189
00:08:14,490 --> 00:08:16,580
you could decompose
your encapsulant

190
00:08:16,580 --> 00:08:20,252
and cause a degradation of not
only the encapsulant itself,

191
00:08:20,252 --> 00:08:21,960
which would block some
of the light going

192
00:08:21,960 --> 00:08:25,560
through, but of the solar
cell absorber material, too.

193
00:08:25,560 --> 00:08:28,990
So that's why, in many
thin film materials,

194
00:08:28,990 --> 00:08:33,480
there are other encapsulants
used, like PVB, polyvinyl

195
00:08:33,480 --> 00:08:34,730
butyral.

196
00:08:34,730 --> 00:08:37,190
And some folks are even
talking about getting away

197
00:08:37,190 --> 00:08:40,020
from glass and EVA
altogether and just

198
00:08:40,020 --> 00:08:44,230
putting down, say, an
organic resin-type material.

199
00:08:44,230 --> 00:08:48,580
Maybe a very hard
polymeric material

200
00:08:48,580 --> 00:08:50,990
that could be used in a floor
in a clean room, let's say.

201
00:08:50,990 --> 00:08:51,490
Yeah.

202
00:08:51,490 --> 00:08:55,048
AUDIENCE: Is the low iron
content of the glass also

203
00:08:55,048 --> 00:08:58,380
important to prevent
all impurities

204
00:08:58,380 --> 00:09:00,760
getting into the silicon
or are we past that point?

205
00:09:00,760 --> 00:09:02,800
PROFESSOR: So the question
was is the low iron

206
00:09:02,800 --> 00:09:05,460
content in the glass also
useful for preventing impurities

207
00:09:05,460 --> 00:09:06,690
from getting in.

208
00:09:06,690 --> 00:09:10,390
I would say, if you're
producing-- the diffusivity

209
00:09:10,390 --> 00:09:14,060
and solubility of
impurities in a solid

210
00:09:14,060 --> 00:09:15,770
typically follow
Boltzmann statistics,

211
00:09:15,770 --> 00:09:18,040
meaning they increase
exponentially with temperature.

212
00:09:18,040 --> 00:09:19,870
It's an entropy-driven effect.

213
00:09:19,870 --> 00:09:23,910
And as a result of being at room
temperature, the transmission,

214
00:09:23,910 --> 00:09:26,080
or the diffusivity and,
hence, the solubility

215
00:09:26,080 --> 00:09:28,760
of iron inside of the
bulk silicon is very low.

216
00:09:28,760 --> 00:09:32,150
And, even over a 20-year span,
the total diffusion length

217
00:09:32,150 --> 00:09:34,180
of the iron from the
glass into the silicon

218
00:09:34,180 --> 00:09:35,290
would be rather small.

219
00:09:35,290 --> 00:09:38,970
Now, point taken, if you
were to, say, create a quartz

220
00:09:38,970 --> 00:09:41,540
tube to do your phosphorus
diffusion at 800 degrees C

221
00:09:41,540 --> 00:09:45,080
or 850 degrees C and you had
iron inside of that quartz

222
00:09:45,080 --> 00:09:47,540
tube, then it could very
easily diffuse into your wafer.

223
00:09:47,540 --> 00:09:50,780
So it is important to be
considerate of impurities

224
00:09:50,780 --> 00:09:53,870
and the effects thereof but
at low temperatures-- close

225
00:09:53,870 --> 00:09:56,262
to room temperature-- there's
very little risk of it

226
00:09:56,262 --> 00:09:57,440
moving around.

227
00:09:57,440 --> 00:09:57,940
Good.

228
00:09:57,940 --> 00:10:01,750
So Tedlar, this back
skin material right here,

229
00:10:01,750 --> 00:10:04,767
forms an almost
impenetrable back layer.

230
00:10:04,767 --> 00:10:06,350
You might want to
put little asterisks

231
00:10:06,350 --> 00:10:08,930
around-- sorry, little
quotes around impenetrable.

232
00:10:08,930 --> 00:10:10,940
Nothing is absolutely
impenetrable

233
00:10:10,940 --> 00:10:15,410
but it is a very
tough, tough material.

234
00:10:15,410 --> 00:10:18,810
If you were to take a piece
of Tedlar and try to break it,

235
00:10:18,810 --> 00:10:22,250
try to rip it or tear it, even
the strongest in the class

236
00:10:22,250 --> 00:10:24,400
would be challenged
in that regard.

237
00:10:24,400 --> 00:10:27,230
And we'll have a chance
to do that at Fraunhofer.

238
00:10:27,230 --> 00:10:29,640
The aluminum frame
provides rigidity.

239
00:10:29,640 --> 00:10:30,140
Right?

240
00:10:30,140 --> 00:10:32,520
So the mechanical
engineers in the room

241
00:10:32,520 --> 00:10:34,590
would understand
that the bending mode

242
00:10:34,590 --> 00:10:37,820
of this glass, or the
twist, is prevented

243
00:10:37,820 --> 00:10:43,610
by these rigid extruded
aluminum components

244
00:10:43,610 --> 00:10:45,470
right here in the back.

245
00:10:45,470 --> 00:10:48,220
And you can see that
a lot of the mass

246
00:10:48,220 --> 00:10:53,460
is concentrated away
from the centroid, which

247
00:10:53,460 --> 00:10:57,330
results in a larger stiffness,
prevention of a larger bending

248
00:10:57,330 --> 00:10:59,510
moment-- second
moment, if you will.

249
00:10:59,510 --> 00:11:00,010
OK.

250
00:11:00,010 --> 00:11:02,540
So there's a fair amount
of mechanical engineering

251
00:11:02,540 --> 00:11:05,120
that goes into the design
because you want to minimize

252
00:11:05,120 --> 00:11:06,450
commodity material costs.

253
00:11:06,450 --> 00:11:07,850
You want to minimize
the amount of materials

254
00:11:07,850 --> 00:11:09,440
that go into the
module, but you still

255
00:11:09,440 --> 00:11:11,060
want it to last for 20 years.

256
00:11:11,060 --> 00:11:13,135
You still want it to be
resistant to snow loads,

257
00:11:13,135 --> 00:11:13,912
to wind loads.

258
00:11:13,912 --> 00:11:15,370
You don't want it
to break, and you

259
00:11:15,370 --> 00:11:17,411
don't want to make good
on your 20-year warranty.

260
00:11:17,411 --> 00:11:21,180
That costs you money
as a manufacturer.

261
00:11:21,180 --> 00:11:23,530
So the circuit design
in most modules

262
00:11:23,530 --> 00:11:26,300
follows something
very similar to this.

263
00:11:26,300 --> 00:11:28,285
Now, what you see when
you glance at this,

264
00:11:28,285 --> 00:11:30,660
what you should be looking at
is-- here are the contacts.

265
00:11:30,660 --> 00:11:33,590
Those are the leads
coming out of the module.

266
00:11:33,590 --> 00:11:35,940
They're essentially
these leads coming out

267
00:11:35,940 --> 00:11:37,360
in the back of the junction box.

268
00:11:37,360 --> 00:11:40,550
If you want to turn it
around again, here we go.

269
00:11:40,550 --> 00:11:43,210
So here are the module
leads coming out.

270
00:11:43,210 --> 00:11:45,460
And the question is,
how do I string together

271
00:11:45,460 --> 00:11:47,340
the cells inside of my module?

272
00:11:47,340 --> 00:11:49,530
How do I string together
these cells right in here

273
00:11:49,530 --> 00:11:51,840
to provide the greatest
value to the customer?

274
00:11:51,840 --> 00:11:54,020
Do I connect them all in
series and make just one

275
00:11:54,020 --> 00:11:56,354
long, snaky electronic path?

276
00:11:56,354 --> 00:11:57,770
Do I connect them
all in parallel?

277
00:11:57,770 --> 00:12:00,920
But that would require
a lot of wiring.

278
00:12:00,920 --> 00:12:02,530
Do I do something in between?

279
00:12:02,530 --> 00:12:04,530
And if you were to
come up close to here

280
00:12:04,530 --> 00:12:06,330
and inspect it,
what you would see

281
00:12:06,330 --> 00:12:09,370
is that you have
strings in parallel.

282
00:12:09,370 --> 00:12:12,920
So you have one going from
here, down and around.

283
00:12:12,920 --> 00:12:16,106
That's one, two, and three.

284
00:12:16,106 --> 00:12:17,042
Right?

285
00:12:17,042 --> 00:12:19,794
Three of these
strings, and they're

286
00:12:19,794 --> 00:12:22,210
connected-- those three strings
are connected in parallel.

287
00:12:22,210 --> 00:12:26,350
Each string is comprised
of several solar cells.

288
00:12:26,350 --> 00:12:29,140
Now, for historical reasons, the
crystalline silicon solar cell

289
00:12:29,140 --> 00:12:31,730
modules have typically had
strings of around 36 cells

290
00:12:31,730 --> 00:12:33,390
connected in series.

291
00:12:33,390 --> 00:12:35,820
What this does is it yields
a maximum power point

292
00:12:35,820 --> 00:12:38,880
voltage somewhere in the
range of 17 to 18 volts.

293
00:12:38,880 --> 00:12:41,270
If you remember the
VOCs of your solar cells

294
00:12:41,270 --> 00:12:45,040
and that maximum power point,
the voltage of your solar cells

295
00:12:45,040 --> 00:12:47,280
are some around 0.5.

296
00:12:47,280 --> 00:12:50,750
And so you take 36 cells,
divide by 2-- or multiply

297
00:12:50,750 --> 00:12:52,110
by 0.5, if you will.

298
00:12:52,110 --> 00:12:55,040
And you're landing
somewhere in this ballpark.

299
00:12:55,040 --> 00:12:59,380
And what that does is it enables
you to hit, even under lower

300
00:12:59,380 --> 00:13:01,600
light conditions,
the overpotential

301
00:13:01,600 --> 00:13:05,070
needed to charge a
rechargeable battery.

302
00:13:05,070 --> 00:13:08,500
So if there's a 12-volt
battery and the voltage output

303
00:13:08,500 --> 00:13:11,260
of the battery is
between, say, 12

304
00:13:11,260 --> 00:13:13,610
and kind of dropping down
to 10 before it starts

305
00:13:13,610 --> 00:13:16,260
using its utility-- when
you try to start your car

306
00:13:16,260 --> 00:13:19,219
and it doesn't really turn over
and you can start your engine,

307
00:13:19,219 --> 00:13:21,260
it's probably because the
voltage in your battery

308
00:13:21,260 --> 00:13:24,690
has dropped to about 10 volts.

309
00:13:24,690 --> 00:13:28,040
The overpotential needed
to charge it, 13-plus,

310
00:13:28,040 --> 00:13:30,190
that's what you're
hitting with the string.

311
00:13:30,190 --> 00:13:32,220
And that's, for
historical reasons,

312
00:13:32,220 --> 00:13:35,680
why we had this number of
cells connected together.

313
00:13:35,680 --> 00:13:38,010
As grid-tied systems
become more common,

314
00:13:38,010 --> 00:13:40,770
though, this constraint
is being reduced.

315
00:13:40,770 --> 00:13:42,686
Over 90% of all of
our solar modules

316
00:13:42,686 --> 00:13:45,310
today are connected to the grid,
including the ones in my home.

317
00:13:45,310 --> 00:13:48,280
So when I'm not there and it's
producing excess electricity,

318
00:13:48,280 --> 00:13:49,820
it's pumping it into the grid.

319
00:13:49,820 --> 00:13:52,440
And maybe we're using
some of my electrons here.

320
00:13:52,440 --> 00:13:54,320
I don't know.

321
00:13:54,320 --> 00:13:57,380
So as grid-tied systems
become more common,

322
00:13:57,380 --> 00:14:00,200
you care less about meeting
the overpotential requirements

323
00:14:00,200 --> 00:14:00,930
for a battery.

324
00:14:00,930 --> 00:14:04,034
You care more about matching
the output of your solar module

325
00:14:04,034 --> 00:14:05,700
with whatever inverter
technology you're

326
00:14:05,700 --> 00:14:09,710
using to convert the
DC power into AC power.

327
00:14:09,710 --> 00:14:13,300
Number two, why do they
spec the modules that size?

328
00:14:13,300 --> 00:14:16,095
You notice that I could
lift it, just barely.

329
00:14:16,095 --> 00:14:17,220
I'm a university professor.

330
00:14:17,220 --> 00:14:19,510
I sit in my chair and
think most of the time.

331
00:14:19,510 --> 00:14:22,680
A person, somebody strong out
there in the field mounting

332
00:14:22,680 --> 00:14:25,477
modules, say if Jessica in the
back were a module installer,

333
00:14:25,477 --> 00:14:28,060
she could lift this no problem
probably with her pinky finger.

334
00:14:28,060 --> 00:14:30,680
And that's why they're
bite-sized, right?

335
00:14:30,680 --> 00:14:33,974
They're a single-person,
liftable, installable modules.

336
00:14:33,974 --> 00:14:35,640
You can't make them
too big, or else you

337
00:14:35,640 --> 00:14:38,510
start needing cranes which, of
course, has its own benefits

338
00:14:38,510 --> 00:14:39,991
for large field installations.

339
00:14:39,991 --> 00:14:41,990
Wouldn't really make sense
with our current mode

340
00:14:41,990 --> 00:14:43,750
of installation on residential.

341
00:14:43,750 --> 00:14:44,910
AUDIENCE: [INAUDIBLE]

342
00:14:44,910 --> 00:14:45,576
PROFESSOR: Yeah.

343
00:14:48,202 --> 00:14:53,040
AUDIENCE: If you had a
battery-- if you were charging

344
00:14:53,040 --> 00:14:55,970
a battery with the module,
you also wouldn't want

345
00:14:55,970 --> 00:14:59,469
to have too many volts, right?

346
00:14:59,469 --> 00:15:01,510
Maybe that's a really
basic electronics question.

347
00:15:01,510 --> 00:15:05,360
PROFESSOR: Yeah, so
the question is, do you

348
00:15:05,360 --> 00:15:08,040
benefit by having
larger overpotential

349
00:15:08,040 --> 00:15:11,100
when recharging a battery?

350
00:15:11,100 --> 00:15:13,580
Well, I'm not a battery
expert, so I'll definitely

351
00:15:13,580 --> 00:15:14,452
caution on this.

352
00:15:14,452 --> 00:15:16,910
I would imagine that there
would be some law of diminishing

353
00:15:16,910 --> 00:15:17,410
returns.

354
00:15:17,410 --> 00:15:19,534
You would be limited in
terms of the recharge rate,

355
00:15:19,534 --> 00:15:21,900
not by the overpotential
but by some other component

356
00:15:21,900 --> 00:15:23,640
like the electrolyte
diffusivity,

357
00:15:23,640 --> 00:15:26,970
let's say, inside of your
system-- as an example.

358
00:15:26,970 --> 00:15:29,417
But I would defer that
to the battery experts.

359
00:15:29,417 --> 00:15:31,000
My impression, though,
is that you get

360
00:15:31,000 --> 00:15:32,208
a law of diminishing returns.

361
00:15:34,630 --> 00:15:36,110
Packing fraction in the modules.

362
00:15:36,110 --> 00:15:38,800
So we talked about the
different cell sizes

363
00:15:38,800 --> 00:15:41,830
if you're processing
discrete cells.

364
00:15:41,830 --> 00:15:45,120
The higher packing fraction
enables lower glass encapsulant

365
00:15:45,120 --> 00:15:46,980
and cost per watt peak.

366
00:15:46,980 --> 00:15:48,900
The lower packing
fraction, you get

367
00:15:48,900 --> 00:15:52,060
to play some games with
optical concentration.

368
00:15:52,060 --> 00:15:55,300
For here, for instance, if you
have space between your cells,

369
00:15:55,300 --> 00:15:57,940
you can have a diffuse
Lambertian scatterer

370
00:15:57,940 --> 00:16:02,160
in the back which will reflect
the light at a different angle,

371
00:16:02,160 --> 00:16:03,140
scattering it.

372
00:16:03,140 --> 00:16:06,020
And that light could be trapped
by total internal reflection

373
00:16:06,020 --> 00:16:09,130
inside of the glass if the
angle reaching the glass

374
00:16:09,130 --> 00:16:11,200
is, say, 15 degrees or greater.

375
00:16:11,200 --> 00:16:13,070
So you can play
some games there.

376
00:16:13,070 --> 00:16:15,690
That's why the Tedlar
back skin being white

377
00:16:15,690 --> 00:16:20,310
is typically used in most solar
modules, so that we reflect

378
00:16:20,310 --> 00:16:22,350
and then scatter and
then trap the light that

379
00:16:22,350 --> 00:16:24,110
comes in in between the cells.

380
00:16:26,890 --> 00:16:32,340
So to put this into
another fancier,

381
00:16:32,340 --> 00:16:35,160
three-dimensional framework-- we
have our interconnected cells.

382
00:16:35,160 --> 00:16:37,330
We have our encapsulant
above and below.

383
00:16:37,330 --> 00:16:40,100
The encapsulant material
is an interesting thing.

384
00:16:40,100 --> 00:16:41,840
It's solid at room
temperature but it

385
00:16:41,840 --> 00:16:45,460
begins to flow at
around 150, 175 degrees

386
00:16:45,460 --> 00:16:50,460
C which means, if
you've ever put

387
00:16:50,460 --> 00:16:53,360
a plastic bottle in the oven
and you've noticed how it begins

388
00:16:53,360 --> 00:16:57,160
to melt and sag and
droop, that's essentially

389
00:16:57,160 --> 00:16:59,442
the type of flow that you
would experience inside

390
00:16:59,442 --> 00:17:00,650
of this encapsulant material.

391
00:17:00,650 --> 00:17:02,240
And it flows around the cells.

392
00:17:02,240 --> 00:17:05,568
And it fuses them together
with the glass on top,

393
00:17:05,568 --> 00:17:07,359
and they stick together
once they cool down

394
00:17:07,359 --> 00:17:09,410
to room temperature.

395
00:17:09,410 --> 00:17:11,970
That forms what's
called a laminate.

396
00:17:11,970 --> 00:17:14,730
Because now you
have the back skin--

397
00:17:14,730 --> 00:17:16,506
the frame isn't
shown right here--

398
00:17:16,506 --> 00:17:17,839
you have the back skin material.

399
00:17:17,839 --> 00:17:19,290
You have the
encapsulant, the cells,

400
00:17:19,290 --> 00:17:20,706
the encapsulant
on the other side,

401
00:17:20,706 --> 00:17:24,130
and your glass on the front
side all together forming

402
00:17:24,130 --> 00:17:26,000
a nice stack of materials.

403
00:17:26,000 --> 00:17:28,610
And they're stuck together,
or bound, by the encapsulant.

404
00:17:28,610 --> 00:17:30,080
That's called a laminate.

405
00:17:30,080 --> 00:17:33,430
It doesn't resist bending
very well, or twist very well,

406
00:17:33,430 --> 00:17:35,280
because it doesn't yet
have the frame on it.

407
00:17:35,280 --> 00:17:37,630
But it certainly can be
transported around the factory

408
00:17:37,630 --> 00:17:39,540
and moved from one
station to another

409
00:17:39,540 --> 00:17:42,290
without risk of, say,
the cells falling out.

410
00:17:42,290 --> 00:17:45,470
They're pretty well fused there.

411
00:17:45,470 --> 00:17:47,390
In terms of moving
to new materials,

412
00:17:47,390 --> 00:17:50,090
there's a lot of work right now
to squeeze the commodity costs

413
00:17:50,090 --> 00:17:52,240
out of solar module
manufacturing.

414
00:17:52,240 --> 00:17:55,260
If you look at the percentage
of cost of commodity materials

415
00:17:55,260 --> 00:17:57,230
in the solar module,
it is quite high.

416
00:17:57,230 --> 00:17:59,580
And so people are focusing
on new types of transparent

417
00:17:59,580 --> 00:18:02,000
front surface materials
that let the light in,

418
00:18:02,000 --> 00:18:03,641
glass replacements.

419
00:18:03,641 --> 00:18:04,140
Right?

420
00:18:04,140 --> 00:18:07,660
You could envision--
we have many examples

421
00:18:07,660 --> 00:18:10,280
of really hard, resistant,
polymeric materials

422
00:18:10,280 --> 00:18:11,300
in our daily lives.

423
00:18:11,300 --> 00:18:12,840
People are looking into those.

424
00:18:12,840 --> 00:18:15,070
The encapsulant materials--
ethyl vinyl acetate

425
00:18:15,070 --> 00:18:18,340
is fairly expensive, and there
aren't that many companies

426
00:18:18,340 --> 00:18:19,600
that manufacture it.

427
00:18:19,600 --> 00:18:22,390
And so people are looking
to options to replace it.

428
00:18:22,390 --> 00:18:25,490
Tedlar even more so,
that back skin material.

429
00:18:25,490 --> 00:18:28,234
Although the challenge there
is a lot harder, a lot tougher.

430
00:18:28,234 --> 00:18:28,900
No pun intended.

431
00:18:28,900 --> 00:18:31,500
It's a really strong material.

432
00:18:31,500 --> 00:18:35,740
And a lot of our modern
chemistry or chemical

433
00:18:35,740 --> 00:18:38,070
engineering goes into
making and designing it.

434
00:18:38,070 --> 00:18:41,160
And then finally finding some
replacement for the frame.

435
00:18:41,160 --> 00:18:43,022
People are talking
about frameless modules.

436
00:18:43,022 --> 00:18:44,730
That means taking off
this aluminum piece

437
00:18:44,730 --> 00:18:47,650
right here on the side and
just leaving the laminate

438
00:18:47,650 --> 00:18:49,530
and somehow providing
the mechanical rigidity

439
00:18:49,530 --> 00:18:52,930
to the laminate and avoiding the
cost of the extruded aluminum

440
00:18:52,930 --> 00:18:54,100
components.

441
00:18:54,100 --> 00:18:57,610
So a lot of work going
in in module design.

442
00:18:57,610 --> 00:19:00,530
This is a picture of folks
mounting the aluminum framing

443
00:19:00,530 --> 00:19:02,070
onto the module.

444
00:19:02,070 --> 00:19:08,530
These are essentially
large steel components

445
00:19:08,530 --> 00:19:10,670
that are pushing the
aluminum into the module

446
00:19:10,670 --> 00:19:14,640
and making sure that there's a
snug fit between the aluminum

447
00:19:14,640 --> 00:19:15,890
and the module itself.

448
00:19:15,890 --> 00:19:18,610
As you can see by the size
and shapes of these cells,

449
00:19:18,610 --> 00:19:21,310
that gives you an indication
of what company that is.

450
00:19:24,360 --> 00:19:27,850
Module technology--
there's some work as well

451
00:19:27,850 --> 00:19:30,420
going into
replacing-- let's see,

452
00:19:30,420 --> 00:19:33,450
here we go-- replacing the
tabbing-stringing step.

453
00:19:33,450 --> 00:19:33,950
Right?

454
00:19:33,950 --> 00:19:36,019
So this is a machine
the tabs and strings

455
00:19:36,019 --> 00:19:37,060
the cells to one another.

456
00:19:37,060 --> 00:19:38,341
We'll see one at Fraunhofer.

457
00:19:38,341 --> 00:19:40,090
It's really impressive
when you see it go,

458
00:19:40,090 --> 00:19:41,631
but it's also quite expensive.

459
00:19:41,631 --> 00:19:43,880
And so if you could get rid
of the tabbing stringing--

460
00:19:43,880 --> 00:19:45,710
imagine just lining
up the little cells

461
00:19:45,710 --> 00:19:47,340
like cookies on a cookie sheet.

462
00:19:47,340 --> 00:19:51,180
And they all are interconnected
by contact with whatever back

463
00:19:51,180 --> 00:19:53,330
skin material is on the
back that is printed,

464
00:19:53,330 --> 00:19:55,160
let's say, with the
circuitry already.

465
00:19:55,160 --> 00:19:57,429
You would wind up in
a situation like this,

466
00:19:57,429 --> 00:19:59,220
where you have your
cells in direct contact

467
00:19:59,220 --> 00:20:01,304
with some printed
circuitry on your back.

468
00:20:01,304 --> 00:20:03,720
Boom, boom, boom, boom, boom--
and you're off and running.

469
00:20:03,720 --> 00:20:08,340
In this particular case, you see
these little white dots here.

470
00:20:08,340 --> 00:20:13,380
The grid pattern on that cell
is a radial spiderweb pattern

471
00:20:13,380 --> 00:20:15,200
that emanates from
that central point,

472
00:20:15,200 --> 00:20:17,730
collects current from
nearby-- essentially,

473
00:20:17,730 --> 00:20:19,540
an area about that
big is collected,

474
00:20:19,540 --> 00:20:21,730
let's say by that little
point right there.

475
00:20:21,730 --> 00:20:24,920
And then there's a hole drilled
through the solar cell where

476
00:20:24,920 --> 00:20:27,110
the metal is wrapping
through to the back.

477
00:20:27,110 --> 00:20:29,960
And that's where it's making
contact with the back side

478
00:20:29,960 --> 00:20:31,300
right there.

479
00:20:31,300 --> 00:20:33,480
And there are other
regions of the solar cell

480
00:20:33,480 --> 00:20:38,130
that are making contact with
the bulk, the base, if you will.

481
00:20:38,130 --> 00:20:40,560
So you have both of your
n- and your p-type contacts

482
00:20:40,560 --> 00:20:42,330
on the back side
of your solar cell.

483
00:20:42,330 --> 00:20:44,050
And that's what's
shown in this drawing.

484
00:20:44,050 --> 00:20:46,570
So there are attempts, as
well, to put all the contacts,

485
00:20:46,570 --> 00:20:48,400
or connect the
contacts, onto the back

486
00:20:48,400 --> 00:20:50,770
to enable this
monolithic manufacturing.

487
00:20:50,770 --> 00:20:51,270
Yes.

488
00:20:51,270 --> 00:20:53,907
AUDIENCE: How does
that not short?

489
00:20:53,907 --> 00:20:54,490
PROFESSOR: OK.

490
00:20:54,490 --> 00:20:56,330
So how is that not a short?

491
00:20:56,330 --> 00:20:59,270
Let me show you how that
would not be a short.

492
00:21:01,870 --> 00:21:04,600
What I'm going to do is attempt
to draw two different cell

493
00:21:04,600 --> 00:21:09,020
designs that could
lead to not shorting.

494
00:21:09,020 --> 00:21:11,040
One cell design that
could lead to not shorting

495
00:21:11,040 --> 00:21:15,600
would be-- let's say this
would be our front surface.

496
00:21:15,600 --> 00:21:17,756
Our base is-- I'm going
to draw it flat here

497
00:21:17,756 --> 00:21:19,380
in the back just for
simplicity, but it

498
00:21:19,380 --> 00:21:21,160
could be textured as well.

499
00:21:21,160 --> 00:21:30,750
I'm going to call this a lightly
doped semiconductor material.

500
00:21:30,750 --> 00:21:33,810
And then I could
have heavily doped--

501
00:21:33,810 --> 00:21:37,960
let's say this would be--
I'm going to exaggerate

502
00:21:37,960 --> 00:21:40,000
here just for effect.

503
00:21:40,000 --> 00:21:44,460
This could be my heavily
n-type doped material.

504
00:21:44,460 --> 00:21:47,000
This could be my heavily
p-type doped material.

505
00:21:47,000 --> 00:21:52,180
So I'm drawing p plus, n plus,
another n plus, another p plus.

506
00:21:52,180 --> 00:21:55,540
So I have charge separation
occurring at these regions

507
00:21:55,540 --> 00:21:58,260
right here and then
contacts on the other side.

508
00:21:58,260 --> 00:21:59,510
This is my electron contact.

509
00:21:59,510 --> 00:22:01,050
This is my hole contact.

510
00:22:01,050 --> 00:22:02,760
This is my electron contact.

511
00:22:02,760 --> 00:22:04,490
This is my hole contact.

512
00:22:04,490 --> 00:22:08,020
And now what's happening
is photoexcited carriers

513
00:22:08,020 --> 00:22:09,490
are being generated in here.

514
00:22:09,490 --> 00:22:12,577
But the field that separates
them is all in the back.

515
00:22:12,577 --> 00:22:14,410
And so now the carriers
are being separated,

516
00:22:14,410 --> 00:22:17,030
electrons being collected by
the darker-shaded, rather,

517
00:22:17,030 --> 00:22:19,270
the shaded areas,
and the holes being

518
00:22:19,270 --> 00:22:21,890
collected by these unshaded
contacts here in the back.

519
00:22:21,890 --> 00:22:24,770
And, if I'm clever about
my contact positioning,

520
00:22:24,770 --> 00:22:27,757
here I have kind of an
interdigitated electron

521
00:22:27,757 --> 00:22:28,590
and hole collectors.

522
00:22:28,590 --> 00:22:29,170
Right?

523
00:22:29,170 --> 00:22:33,430
I'm able to collect both without
having any sort of shunting.

524
00:22:33,430 --> 00:22:36,400
Another design might
be-- let's see.

525
00:22:36,400 --> 00:22:38,710
So this is called an
Interdigitated Back Contact

526
00:22:38,710 --> 00:22:41,326
structure, or IBC.

527
00:22:41,326 --> 00:22:43,660
AUDIENCE: What is
it, [INAUDIBLE]?

528
00:22:43,660 --> 00:22:45,610
PROFESSOR: Actually,
less relevant

529
00:22:45,610 --> 00:22:47,120
in this particular case.

530
00:22:47,120 --> 00:22:50,070
In most interdigitated
back contact cells,

531
00:22:50,070 --> 00:22:54,010
it's actually n-type silicon,
lightly doped n-type silicon,

532
00:22:54,010 --> 00:22:58,190
just because point defects tend
to have smaller capture cross

533
00:22:58,190 --> 00:23:00,190
sections than n-type
silicon and p-type silicon.

534
00:23:00,190 --> 00:23:03,090
So minority carrier
lifetime tends to be larger.

535
00:23:03,090 --> 00:23:06,080
Another cell design, again,
looking at cross section,

536
00:23:06,080 --> 00:23:10,850
you might have-- this
being your solar cell--

537
00:23:10,850 --> 00:23:12,850
you could add texture,
but I'm omitting it just

538
00:23:12,850 --> 00:23:14,670
for simplicity.

539
00:23:14,670 --> 00:23:20,130
Now, let's see, we drill a
hole using a laser right here.

540
00:23:20,130 --> 00:23:22,120
So we have a very thin hole.

541
00:23:22,120 --> 00:23:24,300
Of course, in three
dimensions, there's

542
00:23:24,300 --> 00:23:26,470
mechanical rigidity by
the bulk of the material

543
00:23:26,470 --> 00:23:27,560
in the other dimension.

544
00:23:27,560 --> 00:23:29,184
But we have this hole
drilled through--

545
00:23:29,184 --> 00:23:31,890
and we can make about 18,000
of these holes in about

546
00:23:31,890 --> 00:23:33,820
six seconds with modern
laser technology,

547
00:23:33,820 --> 00:23:38,700
so it's really not that
hard to make this happen.

548
00:23:38,700 --> 00:23:41,190
Although a lot of engineering
effort is put into it,

549
00:23:41,190 --> 00:23:42,267
but it is possible.

550
00:23:42,267 --> 00:23:44,600
And then you could have your
contact metallization, say,

551
00:23:44,600 --> 00:23:47,630
for electrons right here.

552
00:23:50,690 --> 00:23:56,490
You could have the p-n
junction like this.

553
00:23:56,490 --> 00:23:59,340
So this, here, could be
your n plus material.

554
00:23:59,340 --> 00:24:03,430
Again, n plus material, this,
by effect, would also be.

555
00:24:03,430 --> 00:24:06,140
And you can have your
p materials over here.

556
00:24:06,140 --> 00:24:08,200
And you could make
contact to the back right

557
00:24:08,200 --> 00:24:11,720
here and contact to
the back right here.

558
00:24:11,720 --> 00:24:15,810
And as long as there's no metal
short between the electron

559
00:24:15,810 --> 00:24:18,530
and hole-- say, for example, you
generate an electron hole pair

560
00:24:18,530 --> 00:24:20,360
here, the electron
goes up to here.

561
00:24:20,360 --> 00:24:23,600
It goes through the metal to
the back and gets pulled out.

562
00:24:23,600 --> 00:24:26,560
And the hole goes to
this contact over here,

563
00:24:26,560 --> 00:24:29,790
and that's how you
separate the charges.

564
00:24:29,790 --> 00:24:32,740
So this would be an example
of an emitter wrap through

565
00:24:32,740 --> 00:24:36,260
or a metal wrap through
device technology.

566
00:24:36,260 --> 00:24:40,070
Also shown-- MWT,
Metal Wrap Through.

567
00:24:40,070 --> 00:24:42,750
So we have at least two
different device geometries

568
00:24:42,750 --> 00:24:45,410
and many more device
architectures that could fit in

569
00:24:45,410 --> 00:24:47,680
with this future vision.

570
00:24:47,680 --> 00:24:51,930
The problem right now is that
the printed back skin materials

571
00:24:51,930 --> 00:24:55,770
cost about five times more
than the non-printed ones

572
00:24:55,770 --> 00:24:57,780
just because they're
very low scale.

573
00:24:57,780 --> 00:24:59,620
People haven't
investigated it much.

574
00:24:59,620 --> 00:25:01,610
So any time you introduce
a new technology

575
00:25:01,610 --> 00:25:03,360
into the market-- you'll
hear it over and over again

576
00:25:03,360 --> 00:25:05,568
in these innovation and
entrepreneurship classes here

577
00:25:05,568 --> 00:25:07,550
at MIT-- you have to
be significantly better

578
00:25:07,550 --> 00:25:09,050
than the competition.

579
00:25:09,050 --> 00:25:10,740
If you're only a
little bit better,

580
00:25:10,740 --> 00:25:13,537
you face an enormous
barrier to entry

581
00:25:13,537 --> 00:25:15,870
because you have the rest of
the industry pedaling along

582
00:25:15,870 --> 00:25:17,540
on their bicycles
in a huge peloton,

583
00:25:17,540 --> 00:25:19,270
benefiting from each
other's drafting,

584
00:25:19,270 --> 00:25:21,720
and you're on your own bike
facing your own headwinds.

585
00:25:21,720 --> 00:25:22,220
Right?

586
00:25:22,220 --> 00:25:24,310
Trying to build everything
up-- your own factory,

587
00:25:24,310 --> 00:25:25,930
your own expertise.

588
00:25:25,930 --> 00:25:26,810
It's challenging.

589
00:25:26,810 --> 00:25:29,300
But I'm not trying to put
a damper on innovation.

590
00:25:29,300 --> 00:25:32,145
I'm just trying to qualify the
areas of greatest opportunity

591
00:25:32,145 --> 00:25:35,099
and lowest resistance to change.

592
00:25:35,099 --> 00:25:36,140
Go ahead, prove me wrong.

593
00:25:36,140 --> 00:25:36,990
Make a difference.

594
00:25:39,940 --> 00:25:42,984
Without that sort of
headstrong mentality of,

595
00:25:42,984 --> 00:25:44,400
gosh, I know I
have something here

596
00:25:44,400 --> 00:25:46,310
and I'm just going to prove
the world that I have it,

597
00:25:46,310 --> 00:25:48,685
we wouldn't have, for example,
the Pilkington plate glass

598
00:25:48,685 --> 00:25:51,325
process working-- or float
glass process, sorry.

599
00:25:51,325 --> 00:25:53,700
We'd still probably be pouring
and grinding and polishing

600
00:25:53,700 --> 00:25:57,580
our panes of glass and
they'd be very expensive.

601
00:25:57,580 --> 00:26:00,520
That was a little side story,
but it took them about 10 years

602
00:26:00,520 --> 00:26:02,310
to develop the
float glass process

603
00:26:02,310 --> 00:26:05,224
inside of a company
owned by a family,

604
00:26:05,224 --> 00:26:06,890
as opposed to a
publicly traded company.

605
00:26:06,890 --> 00:26:08,525
So they could
continue losing money

606
00:26:08,525 --> 00:26:10,650
over a long period of time
until they perfected it.

607
00:26:10,650 --> 00:26:12,899
And once they got it right,
the whole world copied it,

608
00:26:12,899 --> 00:26:14,870
but it took 10 years
of good investment

609
00:26:14,870 --> 00:26:16,340
for them to get it right.

610
00:26:16,340 --> 00:26:17,230
All right.

611
00:26:17,230 --> 00:26:18,620
Let's move on.

612
00:26:18,620 --> 00:26:21,982
We're going to talk about
module spec sheets here.

613
00:26:21,982 --> 00:26:23,440
When you came into
class today, you

614
00:26:23,440 --> 00:26:26,436
should have picked
up one of these.

615
00:26:26,436 --> 00:26:31,490
This is an example of a First
Solar module spec sheet.

616
00:26:31,490 --> 00:26:33,630
What it does is it hits
you front and center.

617
00:26:33,630 --> 00:26:35,760
The company name-- First Solar.

618
00:26:35,760 --> 00:26:40,790
FS Series 3-- that's the name
of the module, and PV module.

619
00:26:40,790 --> 00:26:43,480
So it gives you the
mechanical description,

620
00:26:43,480 --> 00:26:45,482
the size, the weight.

621
00:26:45,482 --> 00:26:46,940
And then, on the
back, it gives you

622
00:26:46,940 --> 00:26:49,130
all of the technical
parameters that you've wanted.

623
00:26:49,130 --> 00:26:51,790
So on the backside
of the-- we'll

624
00:26:51,790 --> 00:26:54,590
go to the First
Solar spec sheet.

625
00:26:54,590 --> 00:26:55,090
OK.

626
00:26:55,090 --> 00:26:57,230
Well, on your spec
sheet right here, you

627
00:26:57,230 --> 00:26:58,950
can look at some
of the parameters.

628
00:26:58,950 --> 00:27:00,940
There's power at the
maximum power point,

629
00:27:00,940 --> 00:27:04,780
voltage and current at
the maximum power point,

630
00:27:04,780 --> 00:27:07,117
the VOC, the ISC, and so forth.

631
00:27:07,117 --> 00:27:09,200
You have a bunch of different
parameters including

632
00:27:09,200 --> 00:27:10,505
the temperature sensitivity.

633
00:27:10,505 --> 00:27:11,170
Right?

634
00:27:11,170 --> 00:27:15,040
How much does the power degrade
with increasing temperature?

635
00:27:15,040 --> 00:27:17,190
So all of these things
we've studied in class,

636
00:27:17,190 --> 00:27:19,070
and now when you pick up
the spec sheet you're a pro.

637
00:27:19,070 --> 00:27:20,090
You look at it, and
you're like, oh yeah.

638
00:27:20,090 --> 00:27:20,790
I know what that does.

639
00:27:20,790 --> 00:27:21,880
I know what that means.

640
00:27:21,880 --> 00:27:26,130
I know how to model the
energy output of the system.

641
00:27:26,130 --> 00:27:28,740
And that's essentially what the
engineers do when they pick up

642
00:27:28,740 --> 00:27:29,700
the spec sheet.

643
00:27:29,700 --> 00:27:30,327
They read it.

644
00:27:30,327 --> 00:27:32,660
And then they input those
parameters into their computer

645
00:27:32,660 --> 00:27:35,822
models, and they can predict how
these modules will perform out

646
00:27:35,822 --> 00:27:38,770
in the field at
a given location.

647
00:27:38,770 --> 00:27:39,330
OK.

648
00:27:39,330 --> 00:27:41,900
And so you have a variety
of different spec sheets.

649
00:27:41,900 --> 00:27:44,240
These are the modules
that are on my roof.

650
00:27:44,240 --> 00:27:47,740
They're Sharp 187-watt modules.

651
00:27:47,740 --> 00:27:51,690
You have, for example,
right here an example

652
00:27:51,690 --> 00:27:53,500
of an Evergreen Solar module.

653
00:27:53,500 --> 00:27:57,040
Let's spend a minute here
looking a little more deeply.

654
00:27:57,040 --> 00:27:59,600
These modules, these
Evergreen Solar modules,

655
00:27:59,600 --> 00:28:03,540
are the same size, but there are
three different power outputs.

656
00:28:03,540 --> 00:28:05,290
And I think we went
through this already.

657
00:28:05,290 --> 00:28:08,230
We have good cells, medium
cells and bad cells.

658
00:28:08,230 --> 00:28:10,980
And they're binned together
in the different modules

659
00:28:10,980 --> 00:28:12,910
so that they can extract
the maximum price

660
00:28:12,910 --> 00:28:14,260
from the consumer.

661
00:28:14,260 --> 00:28:18,150
So that's why you have the
three different ratings of power

662
00:28:18,150 --> 00:28:19,771
for three different
types of modules

663
00:28:19,771 --> 00:28:22,020
with, essentially, the same
form factor, the same size

664
00:28:22,020 --> 00:28:23,050
and weight.

665
00:28:23,050 --> 00:28:25,244
They're just
different-quality cells.

666
00:28:25,244 --> 00:28:26,180
Yeah.

667
00:28:26,180 --> 00:28:27,230
So that's that.

668
00:28:27,230 --> 00:28:31,700
Temperature
coefficients, so forth.

669
00:28:31,700 --> 00:28:33,000
OK.

670
00:28:33,000 --> 00:28:36,100
One thing to note right
here is, on the back sides

671
00:28:36,100 --> 00:28:38,580
of your modules, you typically
have a little diagram that

672
00:28:38,580 --> 00:28:41,235
shows you how to interconnect
one module to the next.

673
00:28:41,235 --> 00:28:42,610
And those are the
cables that are

674
00:28:42,610 --> 00:28:44,930
used to do the interconnection.

675
00:28:44,930 --> 00:28:47,200
There are some companies
trying to get fancy right now

676
00:28:47,200 --> 00:28:51,340
and saying, well, instead
of interconnecting and then

677
00:28:51,340 --> 00:28:53,580
affixing the modules
separately, why don't we

678
00:28:53,580 --> 00:28:56,340
just make these modules
that kind of click together?

679
00:28:56,340 --> 00:28:59,120
And part of the clicking
process is the interconnection

680
00:28:59,120 --> 00:29:01,050
electrically, and the
other part, obviously,

681
00:29:01,050 --> 00:29:02,620
is the mechanical rigidity.

682
00:29:02,620 --> 00:29:07,920
So there's certainly room for
optimization on module design.

683
00:29:07,920 --> 00:29:08,420
Yeah.

684
00:29:08,420 --> 00:29:09,100
Jessica.

685
00:29:09,100 --> 00:29:10,600
AUDIENCE: When they're
sold-- when modules are sold,

686
00:29:10,600 --> 00:29:13,432
they're advertised for their
voltage at maximum power point,

687
00:29:13,432 --> 00:29:13,931
right?

688
00:29:13,931 --> 00:29:15,095
PROFESSOR: Mm-hmm.

689
00:29:15,095 --> 00:29:17,636
AUDIENCE: What's a software that
we should use-- I'm actually

690
00:29:17,636 --> 00:29:18,046
using this for another class.

691
00:29:18,046 --> 00:29:20,840
What's the software that we
should use that maximum power

692
00:29:20,840 --> 00:29:24,291
point voltage into-- translate
that into what it's actually

693
00:29:24,291 --> 00:29:26,737
going to be? [INAUDIBLE]

694
00:29:26,737 --> 00:29:27,320
PROFESSOR: OK.

695
00:29:27,320 --> 00:29:32,130
So the question became how do
I go from a module spec sheet,

696
00:29:32,130 --> 00:29:35,590
like this, with this
data right here,

697
00:29:35,590 --> 00:29:38,190
and translate that into number
of kilowatt hours produced

698
00:29:38,190 --> 00:29:43,100
per unit time-- day, year--
at a given site location.

699
00:29:43,100 --> 00:29:47,220
The best program that I know
of that encompasses all of this

700
00:29:47,220 --> 00:29:50,700
would be a program called
PVWatts, produced by NREL.

701
00:29:50,700 --> 00:29:54,420
PVWatts, all together,
and you can look it up,

702
00:29:54,420 --> 00:29:56,030
you can use it
online, and you can

703
00:29:56,030 --> 00:30:02,570
input your specific module spec
parameters into the program.

704
00:30:02,570 --> 00:30:05,490
Now, that said, we
have a lot of the tools

705
00:30:05,490 --> 00:30:07,010
to do the estimates ourselves.

706
00:30:07,010 --> 00:30:08,980
If you ruffle back
to homework-- I

707
00:30:08,980 --> 00:30:10,590
think it was
homework number two,

708
00:30:10,590 --> 00:30:13,000
we calculated the
output of a solar system

709
00:30:13,000 --> 00:30:16,270
based on the module
spec parameters.

710
00:30:16,270 --> 00:30:19,070
And so now that we can
pick up a module spec sheet

711
00:30:19,070 --> 00:30:20,470
and read it in a
bit more detail,

712
00:30:20,470 --> 00:30:23,688
we can perform those estimates
a little bit more accurately.

713
00:30:23,688 --> 00:30:24,790
Yep.

714
00:30:24,790 --> 00:30:25,290
Omar.

715
00:30:25,290 --> 00:30:27,755
AUDIENCE: Your first solar
panel is a lot smaller

716
00:30:27,755 --> 00:30:29,400
than that panel right there?

717
00:30:29,400 --> 00:30:30,250
PROFESSOR: Yep.

718
00:30:30,250 --> 00:30:32,630
AUDIENCE: So is there a
reason why First Solar

719
00:30:32,630 --> 00:30:33,630
goes with smaller cells?

720
00:30:33,630 --> 00:30:34,296
PROFESSOR: Yeah.

721
00:30:34,296 --> 00:30:38,530
So why is this First Solar
module, the FS Series 3 module,

722
00:30:38,530 --> 00:30:40,770
a little smaller than some
of the crystalline silicon

723
00:30:40,770 --> 00:30:42,200
modules?

724
00:30:42,200 --> 00:30:44,690
For thin film
deposition, especially

725
00:30:44,690 --> 00:30:47,050
for this closed-space
sublimation process

726
00:30:47,050 --> 00:30:50,080
that First Solar
uses, you're limited

727
00:30:50,080 --> 00:30:52,910
in terms of the deposition area
by the uniformity of deposition

728
00:30:52,910 --> 00:30:55,090
over a large region.

729
00:30:55,090 --> 00:30:56,870
So if you're sputtering
the material,

730
00:30:56,870 --> 00:30:59,440
it's typically
related to the size

731
00:30:59,440 --> 00:31:02,610
of your target and the
distance to your substrate.

732
00:31:02,610 --> 00:31:05,400
That will limit the maximum size
that you can deposit uniformly.

733
00:31:05,400 --> 00:31:08,900
And you probably want to deposit
within plus or minus 2%, let's

734
00:31:08,900 --> 00:31:09,400
say.

735
00:31:09,400 --> 00:31:10,420
Maybe 5%.

736
00:31:10,420 --> 00:31:13,430
I'm not extremely privy to the
precise tolerances in the PV

737
00:31:13,430 --> 00:31:16,230
industry in terms of thin
film thickness variation,

738
00:31:16,230 --> 00:31:19,980
but that would be my estimate.

739
00:31:19,980 --> 00:31:22,570
And in terms of these thermal
evaporation processes,

740
00:31:22,570 --> 00:31:25,430
again, it's how the
machine is designed.

741
00:31:25,430 --> 00:31:28,640
So I imagine you could
probably go bigger.

742
00:31:28,640 --> 00:31:30,060
Is it cost effective?

743
00:31:30,060 --> 00:31:31,270
Question mark.

744
00:31:31,270 --> 00:31:34,950
Somebody at First Solar most
obviously did the cost analysis

745
00:31:34,950 --> 00:31:37,440
and said that this is the
optimum between a variety

746
00:31:37,440 --> 00:31:39,880
of different factors, between
manufacturability and ease

747
00:31:39,880 --> 00:31:42,475
of installation.

748
00:31:42,475 --> 00:31:42,975
Yeah.

749
00:31:42,975 --> 00:31:45,433
AUDIENCE: Is there any reason
why they don't put efficiency

750
00:31:45,433 --> 00:31:46,390
on this sheet?

751
00:31:46,390 --> 00:31:48,030
PROFESSOR: Is there a reason
why they don't put efficiency

752
00:31:48,030 --> 00:31:48,950
on the sheet?

753
00:31:48,950 --> 00:31:50,700
Well, you could
probably very easily

754
00:31:50,700 --> 00:31:53,470
calculate the efficiency, right?

755
00:31:53,470 --> 00:31:56,100
There are two types of
module efficiencies.

756
00:31:56,100 --> 00:31:59,420
There is what's called
total area efficiency, where

757
00:31:59,420 --> 00:32:01,080
you take the total
area of your module

758
00:32:01,080 --> 00:32:02,740
all the way up to
the aluminum frame

759
00:32:02,740 --> 00:32:05,140
and calculate that as your area.

760
00:32:05,140 --> 00:32:07,640
So you take your power--
your maximum power

761
00:32:07,640 --> 00:32:11,140
rated at AM 1.5 sun conditions
and divide it by the area.

762
00:32:11,140 --> 00:32:17,244
There's also a variety of
other less straightforward ways

763
00:32:17,244 --> 00:32:18,910
of calculating
efficiency, some of which

764
00:32:18,910 --> 00:32:21,250
take the active area
of the cells only.

765
00:32:21,250 --> 00:32:23,740
Others exclude the frame.

766
00:32:23,740 --> 00:32:28,570
And so the module efficiency is
not extremely straightforward

767
00:32:28,570 --> 00:32:29,070
parameter.

768
00:32:29,070 --> 00:32:32,817
It has to be specified based
on what areas is assumed.

769
00:32:32,817 --> 00:32:34,400
That would be my
estimate for why they

770
00:32:34,400 --> 00:32:35,566
didn't put it right on here.

771
00:32:35,566 --> 00:32:38,159
And secondly, their module
efficiency, if you do the math,

772
00:32:38,159 --> 00:32:40,450
that one probably turns out
to be somewhere between 11%

773
00:32:40,450 --> 00:32:41,890
and 11.5%.

774
00:32:41,890 --> 00:32:44,740
Now they're hitting 12% or so.

775
00:32:44,740 --> 00:32:46,960
That's not a very high
number compared to, say,

776
00:32:46,960 --> 00:32:48,390
a crystalline silicon module.

777
00:32:48,390 --> 00:32:50,630
And so it wouldn't be
probably something they

778
00:32:50,630 --> 00:32:52,484
advertise in their spec sheet.

779
00:32:56,360 --> 00:32:58,930
Describe how PV module
power output is affected

780
00:32:58,930 --> 00:33:01,510
by the cell mismatch losses.

781
00:33:01,510 --> 00:33:03,890
Warning-- I'm going to
lose some of you on this.

782
00:33:03,890 --> 00:33:07,330
Those of you without a very
strong electrical engineering

783
00:33:07,330 --> 00:33:09,470
or physics background,
I'm going to pick you up

784
00:33:09,470 --> 00:33:13,900
at point number three, so
plant a flag right here.

785
00:33:13,900 --> 00:33:14,860
We'll come back to it.

786
00:33:14,860 --> 00:33:18,220
But for those who
care to follow along,

787
00:33:18,220 --> 00:33:20,290
this is some really
fun stuff that

788
00:33:20,290 --> 00:33:23,180
can get into why we want
to match our cells properly

789
00:33:23,180 --> 00:33:24,840
in terms of their outputs.

790
00:33:24,840 --> 00:33:26,750
So the total current
output-- this

791
00:33:26,750 --> 00:33:29,570
is the ideal diode
equation without all

792
00:33:29,570 --> 00:33:31,700
the fancy two-diode
model, recombination

793
00:33:31,700 --> 00:33:34,330
of the space charge region,
recombination of the bulk,

794
00:33:34,330 --> 00:33:36,550
without the series resistance
and shunt resistance.

795
00:33:36,550 --> 00:33:40,280
Just a very simple explanation
of this ideal diode model.

796
00:33:40,280 --> 00:33:44,130
What we're doing is our M is
the number of cells in parallel

797
00:33:44,130 --> 00:33:46,450
and our N is the number
of cells in series.

798
00:33:46,450 --> 00:33:49,520
So we go back to our
equivalent circuit diagrams

799
00:33:49,520 --> 00:33:52,940
where we have two voltage
current sources that

800
00:33:52,940 --> 00:33:54,460
are connected in series.

801
00:33:54,460 --> 00:33:56,930
If you connect
those two in series,

802
00:33:56,930 --> 00:33:59,460
now the voltages will add.

803
00:33:59,460 --> 00:34:02,500
So the effective voltage
across the entire system,

804
00:34:02,500 --> 00:34:05,360
if you draw a black box around
your two individual voltage

805
00:34:05,360 --> 00:34:07,850
producers, the effective
voltage across the entire system

806
00:34:07,850 --> 00:34:11,900
will be voltage 1 plus
voltage 2, in series.

807
00:34:11,900 --> 00:34:15,280
The currents-- now, if you
think about the current,

808
00:34:15,280 --> 00:34:17,489
you can think of current--
the first order--

809
00:34:17,489 --> 00:34:21,941
as a river of electrons trying
to flow through your circuit.

810
00:34:21,941 --> 00:34:23,440
If there's any
bottleneck somewhere,

811
00:34:23,440 --> 00:34:25,590
that's going to limit
the current that can flow

812
00:34:25,590 --> 00:34:27,389
through the entire circuit.

813
00:34:27,389 --> 00:34:29,520
So the combined
current is going to be

814
00:34:29,520 --> 00:34:31,982
limited by the worst performer.

815
00:34:31,982 --> 00:34:34,190
In mathematical terms, it
would be the harmonic mean.

816
00:34:34,190 --> 00:34:36,690
1 over effective current
would be 1 over current 1

817
00:34:36,690 --> 00:34:38,150
plus 1 over current 2.

818
00:34:38,150 --> 00:34:42,469
So what we see here is, if
we have a number of cells

819
00:34:42,469 --> 00:34:45,526
connected in
parallel-- now, instead

820
00:34:45,526 --> 00:34:47,400
of having them connected
in series like this,

821
00:34:47,400 --> 00:34:50,158
you connect them in parallel--
now the currents are adding.

822
00:34:50,158 --> 00:34:52,449
And the voltage will be
limited by the worst performer.

823
00:34:52,449 --> 00:34:54,040
Because the
potential across both

824
00:34:54,040 --> 00:34:55,630
is going to be
limited by whatever

825
00:34:55,630 --> 00:34:57,300
the lowest potential is.

826
00:34:57,300 --> 00:35:01,080
So when we connect in
parallel, we add currents.

827
00:35:01,080 --> 00:35:03,130
And when we connect in
series, we add voltages.

828
00:35:03,130 --> 00:35:05,550
That's a very simple
way to think about it,

829
00:35:05,550 --> 00:35:10,360
and that's why we have the
M appearing here up top.

830
00:35:10,360 --> 00:35:13,022
If you have a number of
cells connected in parallel,

831
00:35:13,022 --> 00:35:14,480
you will essentially
be multiplying

832
00:35:14,480 --> 00:35:18,050
the [? illuminated ?] current
of one cell by whatever M it is.

833
00:35:18,050 --> 00:35:20,750
And if you have a number of
cells connected in series,

834
00:35:20,750 --> 00:35:23,490
that will impact the voltage.

835
00:35:23,490 --> 00:35:23,990
OK.

836
00:35:23,990 --> 00:35:31,580
So that's an expression
that is useful for N and M

837
00:35:31,580 --> 00:35:35,360
number of identical cells
connected in either series

838
00:35:35,360 --> 00:35:36,980
or parallel, respectively.

839
00:35:36,980 --> 00:35:41,260
And in practice, if we have
bad performers in the bunch,

840
00:35:41,260 --> 00:35:43,080
if we have one cell
that's performing worse

841
00:35:43,080 --> 00:35:46,320
than another, either because
it is intrinsically worse--

842
00:35:46,320 --> 00:35:49,360
it is a defective cell
that somehow got in there

843
00:35:49,360 --> 00:35:52,990
or became defective during use--
or it's a temporary effect.

844
00:35:52,990 --> 00:35:55,580
Maybe a seagull just
landed on top of that cell

845
00:35:55,580 --> 00:35:56,330
and is shading it.

846
00:35:56,330 --> 00:35:57,650
Right?

847
00:35:57,650 --> 00:36:00,800
So whatever the reason is--
temporary or permanent--

848
00:36:00,800 --> 00:36:05,260
that cell number 2,
now, is a bad apple.

849
00:36:05,260 --> 00:36:09,000
And so if we're looking
at parallel mismatches,

850
00:36:09,000 --> 00:36:11,430
we take what we know
about cells in parallel.

851
00:36:11,430 --> 00:36:12,150
OK.

852
00:36:12,150 --> 00:36:15,110
The voltage is limited
by the worst performer.

853
00:36:15,110 --> 00:36:17,242
And we look at the
individual IV curves

854
00:36:17,242 --> 00:36:18,450
of the good and the bad cell.

855
00:36:18,450 --> 00:36:23,170
This is the good cell right
here, this curve there.

856
00:36:23,170 --> 00:36:23,680
Right?

857
00:36:23,680 --> 00:36:26,412
And so this, now,
is positive power.

858
00:36:26,412 --> 00:36:28,370
Essentially, power coming
out of the solar cell

859
00:36:28,370 --> 00:36:30,040
is plotted in the
first quadrant.

860
00:36:30,040 --> 00:36:32,180
We have a lot of
power coming out

861
00:36:32,180 --> 00:36:34,440
and a large voltage
for our good cell.

862
00:36:34,440 --> 00:36:37,050
Here's our bad cell right there.

863
00:36:37,050 --> 00:36:39,560
So slightly lower
current outputs

864
00:36:39,560 --> 00:36:41,640
given by the logarithmic effect.

865
00:36:41,640 --> 00:36:44,980
And the voltage is lower,
significantly lower,

866
00:36:44,980 --> 00:36:46,320
than, say, the good cell.

867
00:36:46,320 --> 00:36:48,060
And so the combined
output right here,

868
00:36:48,060 --> 00:36:50,100
the current at short
circuit current conditions,

869
00:36:50,100 --> 00:36:52,220
is going to be the sum
of the good and the bad.

870
00:36:52,220 --> 00:36:53,660
So that's there.

871
00:36:53,660 --> 00:36:57,050
The voltage is going to be
the harmonic mean, so limited

872
00:36:57,050 --> 00:36:59,990
by the worst performer.

873
00:36:59,990 --> 00:37:04,420
That's how low-voltage
cells can really

874
00:37:04,420 --> 00:37:06,170
mess up parallel strings.

875
00:37:06,170 --> 00:37:11,560
So for example, this set
of two rows of cells,

876
00:37:11,560 --> 00:37:13,940
this one right here, which
is all connected in series.

877
00:37:13,940 --> 00:37:17,650
It's connected in parallel
to this one and this one.

878
00:37:17,650 --> 00:37:21,960
And so if one of these
three subsets of cells

879
00:37:21,960 --> 00:37:24,710
has a lower voltage
than the others,

880
00:37:24,710 --> 00:37:27,280
the combined voltage
output of the entire module

881
00:37:27,280 --> 00:37:28,960
will be lower.

882
00:37:28,960 --> 00:37:32,350
Similarly, now we've
put them in series,

883
00:37:32,350 --> 00:37:34,420
if we have a bad
cell that's producing

884
00:37:34,420 --> 00:37:37,050
less current than the
good cell-- and notice

885
00:37:37,050 --> 00:37:38,900
that they're saying
open circuit voltage now

886
00:37:38,900 --> 00:37:47,330
just for the thought experiment
here in the classroom.

887
00:37:47,330 --> 00:37:49,870
Typically, if a cell is
performing worse than current,

888
00:37:49,870 --> 00:37:50,860
it will also have
a lower voltage.

889
00:37:50,860 --> 00:37:52,818
But just for the thought
experiment right here,

890
00:37:52,818 --> 00:37:56,640
we have bad cell, good cell.

891
00:37:56,640 --> 00:37:58,880
Bad cell has about half the
current of the good cell.

892
00:37:58,880 --> 00:38:01,060
So the voltage output
in series is going

893
00:38:01,060 --> 00:38:02,710
to be the addition of the two.

894
00:38:02,710 --> 00:38:03,380
Boom, boom.

895
00:38:03,380 --> 00:38:04,440
Then you're up to here.

896
00:38:04,440 --> 00:38:07,190
But the current output will be
limited by the worst performer.

897
00:38:07,190 --> 00:38:10,140
And again, the combined
output of this set

898
00:38:10,140 --> 00:38:13,860
of cells connected in series is
going to be right around here.

899
00:38:13,860 --> 00:38:16,070
And so based on your
cell performance--

900
00:38:16,070 --> 00:38:19,310
if you know that
your voltage has

901
00:38:19,310 --> 00:38:21,450
a certain variance in
your manufacturing line

902
00:38:21,450 --> 00:38:23,991
and your current has a certain
variance in your manufacturing

903
00:38:23,991 --> 00:38:26,190
line, that might entice
you to string together

904
00:38:26,190 --> 00:38:28,940
your cells in one
way or another,

905
00:38:28,940 --> 00:38:32,880
depending on how you want
to mitigate your losses.

906
00:38:32,880 --> 00:38:36,060
Shaded cells-- so up to
now, everybody kind of

907
00:38:36,060 --> 00:38:39,260
gets based on Kirchhoff's
laws and so forth

908
00:38:39,260 --> 00:38:42,064
from equivalent circuit
diagrams, so we're rolling.

909
00:38:42,064 --> 00:38:43,730
Now we're going to
get the shaded cells.

910
00:38:43,730 --> 00:38:46,630
And here's where it
becomes a little bit more

911
00:38:46,630 --> 00:38:48,660
difficult to understand
absent the reading

912
00:38:48,660 --> 00:38:49,660
in that book over there.

913
00:38:49,660 --> 00:38:51,440
Did everybody get a chance
to see the blue book?

914
00:38:51,440 --> 00:38:52,920
Did it make its
rounds over here?

915
00:38:52,920 --> 00:38:53,650
No, it didn't.

916
00:38:53,650 --> 00:38:56,190
Why don't we pass
the blue book around?

917
00:38:56,190 --> 00:38:58,750
Omar over there didn't get
a chance to see it yet.

918
00:38:58,750 --> 00:39:00,250
We're going to
consider that we have

919
00:39:00,250 --> 00:39:03,250
10 identical solar cells
connected in series,

920
00:39:03,250 --> 00:39:05,340
and this one over
here is shaded.

921
00:39:05,340 --> 00:39:07,740
What happens now?

922
00:39:07,740 --> 00:39:11,590
So this is described, I believe,
in figure 5.7 in the book, one

923
00:39:11,590 --> 00:39:13,600
with the green little tab.

924
00:39:13,600 --> 00:39:17,610
We've assumed that the
combined IV characteristic

925
00:39:17,610 --> 00:39:21,500
of the n good cells
is this one right here

926
00:39:21,500 --> 00:39:25,270
and the one bad cell
is this right here.

927
00:39:25,270 --> 00:39:30,810
So it's a little bit of a
strange IV characteristic.

928
00:39:30,810 --> 00:39:33,450
And what happens next?

929
00:39:33,450 --> 00:39:38,050
If you combine the two in series
and you use the same protocol

930
00:39:38,050 --> 00:39:41,000
that we've done so far,
the voltage will add.

931
00:39:41,000 --> 00:39:44,780
So whatever voltage gain right
here has been added over here.

932
00:39:44,780 --> 00:39:46,930
But the current is
going to be limited

933
00:39:46,930 --> 00:39:48,830
by the worst performer,
which happens to be

934
00:39:48,830 --> 00:39:50,800
this solar cell right here.

935
00:39:50,800 --> 00:39:52,640
And so the combined
output is going

936
00:39:52,640 --> 00:39:56,280
to be lower-- significantly
lower-- than those

937
00:39:56,280 --> 00:39:57,030
of the good cells.

938
00:39:57,030 --> 00:39:59,910
And that's what happens
when one cell in a string

939
00:39:59,910 --> 00:40:02,800
is shaded if there aren't any
fancy electronics to allow

940
00:40:02,800 --> 00:40:05,000
the current to bypass it.

941
00:40:05,000 --> 00:40:07,710
Furthermore, this
amount of power

942
00:40:07,710 --> 00:40:09,840
will be dissipated
in the bad cell when

943
00:40:09,840 --> 00:40:11,990
the string is short circuited.

944
00:40:11,990 --> 00:40:15,080
It can change a bit when you
start moving toward operating

945
00:40:15,080 --> 00:40:17,940
conditions, but it's a
sizable amount of power being

946
00:40:17,940 --> 00:40:20,370
dumped into that bad cell.

947
00:40:20,370 --> 00:40:23,820
And in practice, that
cell is reverse biased.

948
00:40:23,820 --> 00:40:26,490
Let me show you using
our wonderful PV

949
00:40:26,490 --> 00:40:32,900
CD-ROM exactly why that is.

950
00:40:32,900 --> 00:40:37,090
So here we have two solar
cells connected in series.

951
00:40:37,090 --> 00:40:41,840
And they're both producing
an equal amount of current

952
00:40:41,840 --> 00:40:44,880
because they have an equal
number of these yellow arrows

953
00:40:44,880 --> 00:40:46,380
incident on them.

954
00:40:46,380 --> 00:40:48,190
And they're equal in
every other aspect.

955
00:40:48,190 --> 00:40:51,250
So it's still a thought
experiment right now.

956
00:40:51,250 --> 00:40:54,560
The current that is cycling
through this circuit,

957
00:40:54,560 --> 00:40:57,320
in the external
circuit right here--

958
00:40:57,320 --> 00:41:00,272
the magnitude of that current
is being indicated by the color.

959
00:41:00,272 --> 00:41:02,230
And, in this particular
case, it's very bright.

960
00:41:02,230 --> 00:41:03,040
It's bright green.

961
00:41:03,040 --> 00:41:05,940
It indicates a lot of current
is flowing through it.

962
00:41:05,940 --> 00:41:08,040
So, up to now,
everything makes sense.

963
00:41:08,040 --> 00:41:10,050
This is in short
circuit conditions,

964
00:41:10,050 --> 00:41:11,020
so there's no voltage.

965
00:41:11,020 --> 00:41:13,620
There's no potential
across either cell.

966
00:41:13,620 --> 00:41:15,030
It's in short
circuit conditions,

967
00:41:15,030 --> 00:41:18,710
and current is flowing through
this circuit right here.

968
00:41:18,710 --> 00:41:20,640
Now I'm going to
mismatch the two cells.

969
00:41:20,640 --> 00:41:25,381
I'm going to exchange one of
the cells for a bad apple.

970
00:41:25,381 --> 00:41:25,880
OK.

971
00:41:25,880 --> 00:41:28,082
First off, I know,
just from what

972
00:41:28,082 --> 00:41:30,540
we've been talking about so
far that the current is limited

973
00:41:30,540 --> 00:41:32,540
by the worst performer,
that the current flowing

974
00:41:32,540 --> 00:41:34,910
through the external circuit
is now going to be lower.

975
00:41:34,910 --> 00:41:38,040
And the first order-- it's more
or less approximate to what

976
00:41:38,040 --> 00:41:39,795
the current is in this bad cell.

977
00:41:39,795 --> 00:41:41,170
So the bad cell
right here, which

978
00:41:41,170 --> 00:41:44,700
is indicated by the two arrows
now instead of the four,

979
00:41:44,700 --> 00:41:47,452
this shaded cell right here
is producing less current.

980
00:41:47,452 --> 00:41:49,410
And so I know, based on
what we've talked about

981
00:41:49,410 --> 00:41:51,570
so far, that the
combined current output

982
00:41:51,570 --> 00:41:54,160
of the entire circuit is going
to be approximately this one.

983
00:41:54,160 --> 00:41:57,290
So hence you see a
darker color represented,

984
00:41:57,290 --> 00:42:00,372
the same color that's
flowing through the cell.

985
00:42:00,372 --> 00:42:01,830
So what happens to
the large amount

986
00:42:01,830 --> 00:42:04,050
of current that's being
generated inside of this one?

987
00:42:04,050 --> 00:42:06,067
That cell still has
the capacity to produce

988
00:42:06,067 --> 00:42:07,150
a large amount of current.

989
00:42:07,150 --> 00:42:08,720
Where is it going?

990
00:42:08,720 --> 00:42:12,410
Well, some of that current is
going through this cell right

991
00:42:12,410 --> 00:42:14,305
here, forward biasing
it, which would

992
00:42:14,305 --> 00:42:18,460
result in a negative bias, or
reverse bias, of the bad cell.

993
00:42:18,460 --> 00:42:21,750
Because the potential
across this combined system

994
00:42:21,750 --> 00:42:22,780
still has to be zero.

995
00:42:22,780 --> 00:42:24,410
You're in short
circuit conditions,

996
00:42:24,410 --> 00:42:27,100
so the potential here has to be
equal to the potential there.

997
00:42:27,100 --> 00:42:28,975
So what I've effectively
done is I've reverse

998
00:42:28,975 --> 00:42:32,520
biased my bad cell.

999
00:42:32,520 --> 00:42:34,660
And we'll get to why
that's a problem later on.

1000
00:42:34,660 --> 00:42:36,076
Those who are
electrical engineers

1001
00:42:36,076 --> 00:42:38,110
can already begin
conjuring up ideas

1002
00:42:38,110 --> 00:42:40,200
of reverse biased breakdown.

1003
00:42:40,200 --> 00:42:45,140
We'll explain why reverse
bias is bad in a few.

1004
00:42:45,140 --> 00:42:47,590
Now, if you had a series
of these cells connected--

1005
00:42:47,590 --> 00:42:49,500
a bunch of these cells
connected in series--

1006
00:42:49,500 --> 00:42:52,380
and they were all producing
copious amounts of juice,

1007
00:42:52,380 --> 00:42:54,795
and you had one bad
apple in your string,

1008
00:42:54,795 --> 00:42:56,920
you would be dumping a
considerable amount of power

1009
00:42:56,920 --> 00:42:57,690
across that.

1010
00:42:57,690 --> 00:42:59,939
Or you could be dumping a
considerable amount of power

1011
00:42:59,939 --> 00:43:01,590
across that one cell.

1012
00:43:01,590 --> 00:43:04,640
And that's where
this graph right here

1013
00:43:04,640 --> 00:43:07,666
comes from, the power
dissipated within the bad cell.

1014
00:43:07,666 --> 00:43:09,290
And that's a problem
because if there's

1015
00:43:09,290 --> 00:43:12,210
any weakness in the p-n
junction of that bad cell

1016
00:43:12,210 --> 00:43:14,030
and you're reverse
biasing it, you're

1017
00:43:14,030 --> 00:43:15,582
going to be flooding
a lot of current

1018
00:43:15,582 --> 00:43:17,040
through that one
little spot, which

1019
00:43:17,040 --> 00:43:19,085
means that that spot's
going to heat up.

1020
00:43:19,085 --> 00:43:22,660
And it's going to form a hot
spot, which has the potential

1021
00:43:22,660 --> 00:43:23,450
to get very hot.

1022
00:43:23,450 --> 00:43:25,300
And we know that the
encapsulant materials

1023
00:43:25,300 --> 00:43:28,460
flow at around 150
to 175 degrees C,

1024
00:43:28,460 --> 00:43:31,210
and some other material
failures can occur above that.

1025
00:43:31,210 --> 00:43:35,020
So hot spots are bad.

1026
00:43:35,020 --> 00:43:37,370
If you reverse
bias your devices,

1027
00:43:37,370 --> 00:43:39,830
you can enter what's called
a breakdown regime that,

1028
00:43:39,830 --> 00:43:42,230
essentially, is
driven, oftentimes,

1029
00:43:42,230 --> 00:43:45,820
in a real solar cell by isolated
points and your weaknesses

1030
00:43:45,820 --> 00:43:48,600
in the p-n junction where
current is going to flow.

1031
00:43:48,600 --> 00:43:51,240
And that leads to the hot
spot failure I just mentioned.

1032
00:43:51,240 --> 00:43:55,600
Here's thermographic imaging--
not using the fancy camera

1033
00:43:55,600 --> 00:43:58,485
that we have downstairs, using
a much simpler camera out

1034
00:43:58,485 --> 00:43:59,720
in the field, lower cost.

1035
00:43:59,720 --> 00:44:02,440
These are high enough
temperature variances

1036
00:44:02,440 --> 00:44:05,590
to be able to detect using
pretty low-tech technology.

1037
00:44:05,590 --> 00:44:07,315
It's not a small
amount of current

1038
00:44:07,315 --> 00:44:08,440
going through a solar cell.

1039
00:44:08,440 --> 00:44:10,070
These are large, massive
amounts of current

1040
00:44:10,070 --> 00:44:11,920
going through defective
cells in the field.

1041
00:44:11,920 --> 00:44:16,990
And so you can use these
cameras to visualize

1042
00:44:16,990 --> 00:44:19,369
the underperforming
cells inside of modules.

1043
00:44:19,369 --> 00:44:20,410
Now, that's pretty nifty.

1044
00:44:23,860 --> 00:44:24,730
All right.

1045
00:44:24,730 --> 00:44:27,580
So we planted that flag,
and let's come back to it.

1046
00:44:27,580 --> 00:44:29,330
We're going to describe
how microinverters

1047
00:44:29,330 --> 00:44:32,630
and microelectronics can improve
module performance output.

1048
00:44:32,630 --> 00:44:34,880
Even if you didn't follow
all the detailed explanation

1049
00:44:34,880 --> 00:44:36,680
until now, you can
appreciate the fact

1050
00:44:36,680 --> 00:44:39,200
that a bad apple or bad
solar cell, connected

1051
00:44:39,200 --> 00:44:41,660
in series or in parallel
with the rest of them,

1052
00:44:41,660 --> 00:44:45,660
can cause the combined output
of the module to be lower.

1053
00:44:45,660 --> 00:44:48,260
And so we're going to talk
about how microelectronics

1054
00:44:48,260 --> 00:44:51,330
and eventually microinverters
can help resolve that,

1055
00:44:51,330 --> 00:44:55,950
either on a cell level or on a
larger module-to-module level.

1056
00:44:55,950 --> 00:44:58,610
The simple principle of a
bypass diode of a cell is this.

1057
00:44:58,610 --> 00:45:01,990
If the reverse bias
current becomes too large,

1058
00:45:01,990 --> 00:45:05,010
instead of forcing the current
to flow through the bad cell,

1059
00:45:05,010 --> 00:45:08,310
you can force it to flow
around a bypass diode.

1060
00:45:08,310 --> 00:45:11,180
And therefore, it
short circuits--

1061
00:45:11,180 --> 00:45:12,940
it leaves the cell
out of the circuit.

1062
00:45:12,940 --> 00:45:13,700
Right?

1063
00:45:13,700 --> 00:45:15,430
The current is
flowing around it,

1064
00:45:15,430 --> 00:45:18,220
and so you result in less of
a degradation of your module.

1065
00:45:18,220 --> 00:45:22,410
To put it from an equivalent
IV curve perspective,

1066
00:45:22,410 --> 00:45:25,520
instead of adding this IV curve
here with the rest of them,

1067
00:45:25,520 --> 00:45:27,530
you're simply moving
the current around it

1068
00:45:27,530 --> 00:45:29,940
so you would have the n
good cells, and that's it.

1069
00:45:29,940 --> 00:45:31,050
End of story.

1070
00:45:31,050 --> 00:45:34,070
The 10th cell would just be
taken out of your circuit.

1071
00:45:34,070 --> 00:45:36,970
It would be pushed aside.

1072
00:45:36,970 --> 00:45:37,660
OK.

1073
00:45:37,660 --> 00:45:39,595
So on a cell level,
that make sense.

1074
00:45:42,550 --> 00:45:44,620
Let's skip over a couple
of these, I think.

1075
00:45:44,620 --> 00:45:47,100
On a module level, as well,
it can make a lot of sense.

1076
00:45:47,100 --> 00:45:51,770
And here, it really
does because the shading

1077
00:45:51,770 --> 00:45:54,384
losses on a module level
tend to be more severe.

1078
00:45:54,384 --> 00:45:56,550
You can have, for example,
a tree in the wrong place

1079
00:45:56,550 --> 00:45:58,890
or a telephone pole
somewhere, shading

1080
00:45:58,890 --> 00:46:02,680
one particular module in an
array, an array of modules.

1081
00:46:02,680 --> 00:46:06,506
On my roof, for example,
there are 12 Sharp 187s.

1082
00:46:06,506 --> 00:46:10,280
So if you have a series of
these modules up on the roof,

1083
00:46:10,280 --> 00:46:13,240
you can cause the
current to circumvent

1084
00:46:13,240 --> 00:46:15,910
one of the modules that's
underperforming, again,

1085
00:46:15,910 --> 00:46:17,380
using the bypass diodes.

1086
00:46:17,380 --> 00:46:21,660
And those are fairly standard
in most of the junction boxes.

1087
00:46:21,660 --> 00:46:25,930
So inside of here, there is
some very simple electronics

1088
00:46:25,930 --> 00:46:30,010
to prevent the modules
from becoming power sinks

1089
00:46:30,010 --> 00:46:32,210
and perhaps even drawing
power from the grid.

1090
00:46:36,629 --> 00:46:37,129
OK.

1091
00:46:41,900 --> 00:46:46,160
So microinverters--
some people say, well,

1092
00:46:46,160 --> 00:46:50,560
instead of just managing
power on a DC basis

1093
00:46:50,560 --> 00:46:53,010
and either dealing
with the mismatch loss

1094
00:46:53,010 --> 00:46:56,160
or cutting bad apples out
of the circuit altogether,

1095
00:46:56,160 --> 00:46:58,120
let's do something even better.

1096
00:46:58,120 --> 00:47:03,090
Let's try to-- to put this in
childhood education terms--

1097
00:47:03,090 --> 00:47:06,320
let's let every module perform
up to their greatest potential,

1098
00:47:06,320 --> 00:47:07,487
and then everybody is happy.

1099
00:47:07,487 --> 00:47:07,986
Right?

1100
00:47:07,986 --> 00:47:09,830
So we're going to convert
the power directly

1101
00:47:09,830 --> 00:47:12,367
on the back of the module
into AC, alternating current.

1102
00:47:12,367 --> 00:47:14,450
And then we're going to
use electronics from there

1103
00:47:14,450 --> 00:47:17,750
to essentially add
the currents together

1104
00:47:17,750 --> 00:47:20,690
once we have defined
the voltage at 120 volts

1105
00:47:20,690 --> 00:47:22,100
at alternating current.

1106
00:47:22,100 --> 00:47:25,700
And the advantage of that
is that if one module is

1107
00:47:25,700 --> 00:47:27,920
underperforming, instead
of contributing zero,

1108
00:47:27,920 --> 00:47:30,180
it's now contributing
some fraction.

1109
00:47:30,180 --> 00:47:33,450
If the mismatch was small enough
not to trip any of the bypass

1110
00:47:33,450 --> 00:47:35,960
diodes but still degrade the
power of the overall system,

1111
00:47:35,960 --> 00:47:37,610
now you're able to
recover-- you're

1112
00:47:37,610 --> 00:47:39,580
able to allow that good
module to perform up

1113
00:47:39,580 --> 00:47:42,070
to its true potential
instead of being dragged down

1114
00:47:42,070 --> 00:47:43,190
by the others.

1115
00:47:43,190 --> 00:47:44,850
And so the
microinverter companies

1116
00:47:44,850 --> 00:47:48,290
that are proposing, and
actually marketing and selling,

1117
00:47:48,290 --> 00:47:50,500
modules that apply the
electronics directly

1118
00:47:50,500 --> 00:47:52,250
in the back of the
solar panels have

1119
00:47:52,250 --> 00:47:54,310
the advantage of
being able to eke out

1120
00:47:54,310 --> 00:47:57,190
several additional percent of
total power from the array,

1121
00:47:57,190 --> 00:47:58,420
from the system.

1122
00:47:58,420 --> 00:48:01,550
The disadvantage of distributing
these microinverter components

1123
00:48:01,550 --> 00:48:03,587
over every single module is?

1124
00:48:03,587 --> 00:48:06,170
AUDIENCE: You have to make a lot
of them, and they cost money.

1125
00:48:06,170 --> 00:48:06,850
PROFESSOR: You have
to make a lot of them.

1126
00:48:06,850 --> 00:48:07,516
They cost money.

1127
00:48:07,516 --> 00:48:11,160
They're costly, and-- mean
time between failures?

1128
00:48:11,160 --> 00:48:11,700
Increases.

1129
00:48:11,700 --> 00:48:12,200
Right?

1130
00:48:12,200 --> 00:48:12,750
Potentially.

1131
00:48:12,750 --> 00:48:13,940
Potentially.

1132
00:48:13,940 --> 00:48:16,462
So if you have an
inverter-- how many people

1133
00:48:16,462 --> 00:48:18,420
know how long an inverter
lasts before it needs

1134
00:48:18,420 --> 00:48:20,710
to be replaced, typically?

1135
00:48:20,710 --> 00:48:21,430
5 to 10 years.

1136
00:48:21,430 --> 00:48:24,121
Now it's up to, say,
7 to 12, but yes.

1137
00:48:24,121 --> 00:48:25,120
Somewhere in that range.

1138
00:48:25,120 --> 00:48:28,052
Less than the 20- to 30-year
life span of your array.

1139
00:48:28,052 --> 00:48:30,260
So now if you have these
distributed little inverters

1140
00:48:30,260 --> 00:48:33,664
across every solar
panel and they fizz out

1141
00:48:33,664 --> 00:48:35,330
after a certain amount
of time, somebody

1142
00:48:35,330 --> 00:48:36,830
has to go up there
and replace them.

1143
00:48:36,830 --> 00:48:39,750
First, identify where they went
down and then replace them.

1144
00:48:39,750 --> 00:48:42,700
So microinverter longevity
is an important part

1145
00:48:42,700 --> 00:48:45,060
of reducing risk and
marketing inverters.

1146
00:48:45,060 --> 00:48:48,490
So the microinverter community
is helping to push, or improve,

1147
00:48:48,490 --> 00:48:51,644
the quality of inverters over
the entire solar spectrum,

1148
00:48:51,644 --> 00:48:52,560
which is kind of cool.

1149
00:48:52,560 --> 00:48:53,060
Yeah.

1150
00:48:53,060 --> 00:48:54,484
AUDIENCE: What's
the [INAUDIBLE]?

1151
00:48:57,370 --> 00:48:59,645
PROFESSOR: So, my understanding
from Rajeev Ram, who

1152
00:48:59,645 --> 00:49:02,380
is actually down in
ARPA-E right now--

1153
00:49:02,380 --> 00:49:05,290
he's a professor at MIT
who's been assisting

1154
00:49:05,290 --> 00:49:09,390
the DOE over the past
couple of years-- most

1155
00:49:09,390 --> 00:49:12,120
failure in inverters relates
to the power electronics itself

1156
00:49:12,120 --> 00:49:14,160
and the high speed
switchers that are used.

1157
00:49:14,160 --> 00:49:15,700
That's my understanding.

1158
00:49:15,700 --> 00:49:18,170
I could look into
that a little bit more

1159
00:49:18,170 --> 00:49:22,140
and get you more details about
precise inverter failure modes,

1160
00:49:22,140 --> 00:49:25,257
but that is my
rudimentary understanding

1161
00:49:25,257 --> 00:49:26,340
from discussions with him.

1162
00:49:31,090 --> 00:49:32,640
Sorry, my screen
is still cracked.

1163
00:49:32,640 --> 00:49:34,255
It's going to be fixed tonight.

1164
00:49:34,255 --> 00:49:36,005
Advantages of integrated
power electronics

1165
00:49:36,005 --> 00:49:37,338
is maximum power point tracking.

1166
00:49:37,338 --> 00:49:38,340
This advantage is-- OK.

1167
00:49:38,340 --> 00:49:38,680
Yep.

1168
00:49:38,680 --> 00:49:39,270
OK.

1169
00:49:39,270 --> 00:49:41,830
And so it's still a
relatively hot topic.

1170
00:49:41,830 --> 00:49:46,380
There was an ARPA-E call for
improved power electronics

1171
00:49:46,380 --> 00:49:50,140
across solar and
other industries.

1172
00:49:50,140 --> 00:49:52,514
And there is some discussion
in the PV community

1173
00:49:52,514 --> 00:49:54,805
as to how much to adopt this
new technology, especially

1174
00:49:54,805 --> 00:49:56,910
in residential systems.

1175
00:49:56,910 --> 00:50:00,540
So trends in
installations in general.

1176
00:50:00,540 --> 00:50:06,750
In the past, installations
were heavily labor-intensive.

1177
00:50:06,750 --> 00:50:09,070
There is now a large
amount of pre-assembly

1178
00:50:09,070 --> 00:50:12,190
done before the
modules hit the roof,

1179
00:50:12,190 --> 00:50:14,850
but there's still
a sizable number

1180
00:50:14,850 --> 00:50:18,440
of components that go into
every single solar module.

1181
00:50:18,440 --> 00:50:21,220
And that takes time, to
mount everything out there

1182
00:50:21,220 --> 00:50:22,340
in the field.

1183
00:50:22,340 --> 00:50:25,530
So what are some other--
what is the dream

1184
00:50:25,530 --> 00:50:27,450
of any solar installer?

1185
00:50:27,450 --> 00:50:31,077
Is to get this roll of
material just unrolls, unfurls,

1186
00:50:31,077 --> 00:50:33,660
onto the roof and using a staple
gun-- bang, bang, bang, bang,

1187
00:50:33,660 --> 00:50:35,180
bang-- or a nail gun.

1188
00:50:35,180 --> 00:50:36,780
Everything is locked down.

1189
00:50:36,780 --> 00:50:37,720
It's lightweight.

1190
00:50:37,720 --> 00:50:43,435
It doesn't underperform
when it gets too warm.

1191
00:50:43,435 --> 00:50:45,560
These are all the specs
that would constitute, say,

1192
00:50:45,560 --> 00:50:47,390
an ideal solar module.

1193
00:50:47,390 --> 00:50:48,900
We don't get that lucky.

1194
00:50:48,900 --> 00:50:51,934
We have to make
compromises in reality.

1195
00:50:51,934 --> 00:50:53,350
Say, for example,
on my roof there

1196
00:50:53,350 --> 00:50:55,986
was a soft spot on the roof
that had to be worked around.

1197
00:50:55,986 --> 00:50:57,860
The structural engineer
had to be called back

1198
00:50:57,860 --> 00:51:00,230
and evaluate and figure out
whether or not it needed

1199
00:51:00,230 --> 00:51:01,510
reinforcements and so forth.

1200
00:51:01,510 --> 00:51:05,180
So we're still far
away in terms of where

1201
00:51:05,180 --> 00:51:07,810
we need to be for
solar installations

1202
00:51:07,810 --> 00:51:10,510
to really be truly low cost.

1203
00:51:10,510 --> 00:51:14,510
And a lot of it has to do with
the balance of system design.

1204
00:51:14,510 --> 00:51:17,860
So I mentioned a few
companies are attempting

1205
00:51:17,860 --> 00:51:24,470
to integrate the connections
into this facile click-lock

1206
00:51:24,470 --> 00:51:25,730
mechanism.

1207
00:51:25,730 --> 00:51:28,480
There have been patents filed, I
think, since the 1990s on this.

1208
00:51:28,480 --> 00:51:30,140
Probably even earlier.

1209
00:51:30,140 --> 00:51:32,855
More recently, some
companies have marketed

1210
00:51:32,855 --> 00:51:34,770
a product in this regard.

1211
00:51:34,770 --> 00:51:38,930
And of course, that increases
the facility of installation.

1212
00:51:38,930 --> 00:51:40,660
There's even a few
that are selling,

1213
00:51:40,660 --> 00:51:43,380
I think, it was Akeena at
Lowe's and, for a time,

1214
00:51:43,380 --> 00:51:44,985
BP Solar was selling
at Home Depot.

1215
00:51:44,985 --> 00:51:46,610
They're examples of
companies beginning

1216
00:51:46,610 --> 00:51:49,390
to market their product
at hardware stores.

1217
00:51:49,390 --> 00:51:49,890
Question.

1218
00:51:49,890 --> 00:51:51,390
AUDIENCE: Could you
define exactly what

1219
00:51:51,390 --> 00:51:52,590
you mean by balance of systems?

1220
00:51:52,590 --> 00:51:53,215
PROFESSOR: Yes.

1221
00:51:53,215 --> 00:51:56,640
So balance of systems
is defined as everything

1222
00:51:56,640 --> 00:51:58,300
beyond the module.

1223
00:51:58,300 --> 00:52:00,620
So that would include
the wiring, the framing,

1224
00:52:00,620 --> 00:52:03,860
the racking, the inverter.

1225
00:52:03,860 --> 00:52:04,490
Yes.

1226
00:52:04,490 --> 00:52:06,320
That is balance of systems.

1227
00:52:06,320 --> 00:52:07,900
Excellent question.

1228
00:52:07,900 --> 00:52:10,180
Excellent question.

1229
00:52:10,180 --> 00:52:13,540
So we're going to talk
about some of the tests

1230
00:52:13,540 --> 00:52:15,170
that the PV module
must pass to ensure

1231
00:52:15,170 --> 00:52:17,560
reliable multi-decade
service life in the field

1232
00:52:17,560 --> 00:52:19,640
and some of the
shortcomings of the tests.

1233
00:52:19,640 --> 00:52:22,550
I'm going to motivate
this by saying

1234
00:52:22,550 --> 00:52:26,300
there's reason to be
concerned, in the early days

1235
00:52:26,300 --> 00:52:28,800
of the industry, about
performance and reliability.

1236
00:52:28,800 --> 00:52:31,180
This was a study done
by Arne Jacobson.

1237
00:52:31,180 --> 00:52:33,290
The group that's working
with the World Bank

1238
00:52:33,290 --> 00:52:36,740
will be as well working with
Arne Jacobson on the class

1239
00:52:36,740 --> 00:52:37,560
project.

1240
00:52:37,560 --> 00:52:40,160
He's a professor at Humboldt
University in California.

1241
00:52:40,160 --> 00:52:42,950
During his PhD with
Dan Kammen in Berkeley,

1242
00:52:42,950 --> 00:52:47,180
he did an audit of rated
versus actual power

1243
00:52:47,180 --> 00:52:51,020
outputs of solar modules in
the field of Kenya in Africa

1244
00:52:51,020 --> 00:52:53,390
and found a rather
large discrepancy

1245
00:52:53,390 --> 00:52:56,540
between what was promised and
what was actually delivered.

1246
00:52:56,540 --> 00:53:00,560
This is a great example
of field work being

1247
00:53:00,560 --> 00:53:03,620
done-- good old-fashioned
detective work-- that leads

1248
00:53:03,620 --> 00:53:05,570
to changes in the industry.

1249
00:53:05,570 --> 00:53:07,700
That got a number of
the under-performers

1250
00:53:07,700 --> 00:53:10,360
to improve their product.

1251
00:53:10,360 --> 00:53:14,470
So I mentioned there's a
question of the customer.

1252
00:53:14,470 --> 00:53:17,380
Am I getting the real value
that I've been promised?

1253
00:53:17,380 --> 00:53:19,659
Especially in the absence
of independent monitoring.

1254
00:53:19,659 --> 00:53:21,700
If you're not actively
measuring the power output

1255
00:53:21,700 --> 00:53:24,160
and recording it, there's
always that question lingering.

1256
00:53:24,160 --> 00:53:25,950
So module testing
and reliability

1257
00:53:25,950 --> 00:53:29,880
is meant to ensure product
quality and output quality.

1258
00:53:29,880 --> 00:53:30,885
Secondly, it's safety.

1259
00:53:30,885 --> 00:53:33,010
I mentioned that these
thermographic cameras-- this

1260
00:53:33,010 --> 00:53:35,575
is an example in an
array, of a hot spot.

1261
00:53:35,575 --> 00:53:36,810
It's not too hot right now.

1262
00:53:36,810 --> 00:53:39,180
It's 36 degrees or so,
but everything else

1263
00:53:39,180 --> 00:53:40,690
is down around 15.

1264
00:53:40,690 --> 00:53:43,940
If you imagine this being
on a sunny day-- a really

1265
00:53:43,940 --> 00:53:46,190
sunny day-- this
hot spot potentially

1266
00:53:46,190 --> 00:53:48,470
could be getting up much higher.

1267
00:53:48,470 --> 00:53:49,900
So that's obviously of concern.

1268
00:53:49,900 --> 00:53:54,010
And there's as well
hot spots or heat

1269
00:53:54,010 --> 00:53:58,240
that can be generated from
the junction box on the back.

1270
00:53:58,240 --> 00:54:00,680
This contains a fair amount
of electronics, again.

1271
00:54:00,680 --> 00:54:04,570
And if there is some failure of
the electronics inside of here,

1272
00:54:04,570 --> 00:54:07,640
you can have the combined power
output of the entire module

1273
00:54:07,640 --> 00:54:11,300
arcing through a given
component back here causing

1274
00:54:11,300 --> 00:54:12,810
sparks to fly.

1275
00:54:12,810 --> 00:54:16,400
And this is where this Photon
International article from 2009

1276
00:54:16,400 --> 00:54:21,085
comes in where they say
it's extremely important--

1277
00:54:21,085 --> 00:54:22,460
the upshot of the
entire article,

1278
00:54:22,460 --> 00:54:24,190
I gave it to you
in its entirety.

1279
00:54:24,190 --> 00:54:28,860
But the upshot is let's
not play with fire,

1280
00:54:28,860 --> 00:54:31,030
meaning let's make
sure that the safety

1281
00:54:31,030 --> 00:54:34,040
standards of solar modules
are really up to spec

1282
00:54:34,040 --> 00:54:37,730
so that the risk of something
catastrophically bad happening

1283
00:54:37,730 --> 00:54:39,910
is negligible.

1284
00:54:39,910 --> 00:54:42,610
Because there's nothing
worse for a nascent industry

1285
00:54:42,610 --> 00:54:45,565
than to have some,
dare I say it,

1286
00:54:45,565 --> 00:54:49,640
Fukushima equivalent occur on
some very high-profile event

1287
00:54:49,640 --> 00:54:52,290
that causes public opinion
to sway against it.

1288
00:54:52,290 --> 00:54:55,640
So making sure that
modules are safe

1289
00:54:55,640 --> 00:54:57,730
is of utmost importance
to preserving

1290
00:54:57,730 --> 00:54:59,480
the positive image of solar.

1291
00:54:59,480 --> 00:55:02,190
And that's what this article
here is meant to convey,

1292
00:55:02,190 --> 00:55:04,070
and you're welcome
to read it in detail.

1293
00:55:04,070 --> 00:55:06,950
It cites one particular example
where that didn't happen

1294
00:55:06,950 --> 00:55:13,520
and it did lead to it a little
PR event-- well-justified,

1295
00:55:13,520 --> 00:55:16,240
in fact.

1296
00:55:16,240 --> 00:55:18,570
Shown here are a series
of tests that one

1297
00:55:18,570 --> 00:55:21,820
can conduct with solar modules.

1298
00:55:21,820 --> 00:55:28,460
I showed a list of different IEC
tests, standardized protocols.

1299
00:55:28,460 --> 00:55:33,490
These are a series of tests,
typically between 12 and 20,

1300
00:55:33,490 --> 00:55:37,250
that represent a battery of
tests for the solar module.

1301
00:55:37,250 --> 00:55:40,530
And if they pass it,
they are IEC certified.

1302
00:55:40,530 --> 00:55:42,800
And you'll see on the back
of certain spec sheets,

1303
00:55:42,800 --> 00:55:44,760
like this one from
First Solar right here,

1304
00:55:44,760 --> 00:55:49,770
the different testing agencies
that have certified the module,

1305
00:55:49,770 --> 00:55:52,220
these little logos represented.

1306
00:55:52,220 --> 00:55:56,460
So it says it's UL
certified, CE mark,

1307
00:55:56,460 --> 00:56:00,000
certified according
to IEC spec 61646.

1308
00:56:00,000 --> 00:56:02,250
That's the one right up there
at the top and so forth.

1309
00:56:02,250 --> 00:56:05,730
So certification
is very important.

1310
00:56:05,730 --> 00:56:07,720
What happens during testing?

1311
00:56:07,720 --> 00:56:10,580
Well, we'll see it because when
I designed the Fraunhofer CSE

1312
00:56:10,580 --> 00:56:14,490
labs, we designed it, Roland
Schindler and I, with the IEC

1313
00:56:14,490 --> 00:56:15,480
tests in mind.

1314
00:56:15,480 --> 00:56:17,280
So each of the
pieces of equipment

1315
00:56:17,280 --> 00:56:21,400
out there are to test one
component of the IEC test.

1316
00:56:21,400 --> 00:56:24,860
Included in the test is
environmental cycling,

1317
00:56:24,860 --> 00:56:26,780
where you take a
module, stick it inside

1318
00:56:26,780 --> 00:56:28,027
of what looks like an oven.

1319
00:56:28,027 --> 00:56:29,985
And it heats it up, cools
it down, heats it up,

1320
00:56:29,985 --> 00:56:31,860
cools it down or maintains
a high temperature

1321
00:56:31,860 --> 00:56:32,907
with high humidity.

1322
00:56:32,907 --> 00:56:35,490
Different types of environmental
chambers do different things,

1323
00:56:35,490 --> 00:56:39,540
but they're meant to simulate
an accelerated aging process.

1324
00:56:39,540 --> 00:56:41,840
Again, since most
diffusion processes

1325
00:56:41,840 --> 00:56:44,030
are driven by
Boltzmann statistics,

1326
00:56:44,030 --> 00:56:48,020
there's an exponential increase
with increasing temperature.

1327
00:56:48,020 --> 00:56:51,790
And hence, we can simulate
accelerated degradation

1328
00:56:51,790 --> 00:56:53,290
at higher temperatures.

1329
00:56:53,290 --> 00:56:54,860
So it's not uncommon
that you hear

1330
00:56:54,860 --> 00:56:57,860
about 85/85 tests,
which mean 85 degrees

1331
00:56:57,860 --> 00:57:01,250
Celsius at 85% humidity.

1332
00:57:01,250 --> 00:57:03,020
So that is a massive jungle.

1333
00:57:03,020 --> 00:57:06,110
We don't have, thankfully,
places on earth like that.

1334
00:57:06,110 --> 00:57:09,080
The maximum temperatures
are at around 45, 50,

1335
00:57:09,080 --> 00:57:12,520
55 degrees C typically and,
in those cases, very dry.

1336
00:57:12,520 --> 00:57:14,440
But in here, 85
degrees Celsius--

1337
00:57:14,440 --> 00:57:16,700
it's still below the
boiling point of water.

1338
00:57:16,700 --> 00:57:19,310
Very high humidity
and the humidity,

1339
00:57:19,310 --> 00:57:21,810
if there is any place for the
humidity go inside the module,

1340
00:57:21,810 --> 00:57:24,510
if you have any failure of
your encapsulant material

1341
00:57:24,510 --> 00:57:26,300
or it can get inside, it will.

1342
00:57:26,300 --> 00:57:28,100
And it will begin
degrading your module.

1343
00:57:28,100 --> 00:57:29,680
You'll see the power
output decrease.

1344
00:57:29,680 --> 00:57:33,490
And after three months subjected
to that sort of torture,

1345
00:57:33,490 --> 00:57:36,170
if the module is poorly
designed, oftentimes

1346
00:57:36,170 --> 00:57:37,670
it will fail.

1347
00:57:37,670 --> 00:57:39,590
Now, some of the
shortcomings of these

1348
00:57:39,590 --> 00:57:44,460
tests-- that
humidity cycle isn't

1349
00:57:44,460 --> 00:57:48,900
necessarily, with the module,
experiencing an applied bias

1350
00:57:48,900 --> 00:57:49,580
voltage.

1351
00:57:49,580 --> 00:57:51,590
So if there's any
electromigration affect,

1352
00:57:51,590 --> 00:57:53,370
you might not
detect it with that.

1353
00:57:53,370 --> 00:57:55,350
So there are people
designing new forms

1354
00:57:55,350 --> 00:57:58,312
of tests that will
probe other failure

1355
00:57:58,312 --> 00:58:00,020
modes, other degradation
mechanisms, that

1356
00:58:00,020 --> 00:58:01,978
aren't necessarily captured
in these IEC tests.

1357
00:58:01,978 --> 00:58:02,940
Do you have a question?

1358
00:58:02,940 --> 00:58:03,440
No.

1359
00:58:03,440 --> 00:58:03,950
No question.

1360
00:58:03,950 --> 00:58:04,100
OK.

1361
00:58:04,100 --> 00:58:04,255
Yeah.

1362
00:58:04,255 --> 00:58:04,755
Over here.

1363
00:58:04,755 --> 00:58:07,667
AUDIENCE: What is the failure
mode of degradation by water?

1364
00:58:07,667 --> 00:58:08,500
PROFESSOR: By water?

1365
00:58:08,500 --> 00:58:09,000
Yes.

1366
00:58:09,000 --> 00:58:12,350
So water, humidity--
typically what happens

1367
00:58:12,350 --> 00:58:14,860
is, in the laminate
itself, there

1368
00:58:14,860 --> 00:58:18,617
will be some part at the
edge that water can get in.

1369
00:58:18,617 --> 00:58:20,450
Not always, but in the
poorly designed ones,

1370
00:58:20,450 --> 00:58:22,680
yes, and so the water
can get in and begin

1371
00:58:22,680 --> 00:58:26,170
diffusing into the module.

1372
00:58:26,170 --> 00:58:28,270
So if you have a thin
film material that

1373
00:58:28,270 --> 00:58:31,130
reacts with water, you might
degrade the actual absorber

1374
00:58:31,130 --> 00:58:32,080
layer itself.

1375
00:58:32,080 --> 00:58:34,160
If you have a silicon
device, the water

1376
00:58:34,160 --> 00:58:35,740
could react with
the metal and cause

1377
00:58:35,740 --> 00:58:39,870
it to corrode-- cause it
to oxidize, in other words.

1378
00:58:39,870 --> 00:58:44,340
And of course, in reacting
with the ethyl vinyl acetate,

1379
00:58:44,340 --> 00:58:47,730
it can cause the encapsulant
to delaminate from the cells

1380
00:58:47,730 --> 00:58:48,890
and tear apart.

1381
00:58:48,890 --> 00:58:51,822
And so you see this
discoloration of the module.

1382
00:58:51,822 --> 00:58:53,280
Those are some of
the failure modes

1383
00:58:53,280 --> 00:58:55,830
that can occur when water
gets in through the side.

1384
00:58:55,830 --> 00:58:59,200
If it managed to
permeate the encapsulant

1385
00:58:59,200 --> 00:59:01,480
materials themselves,
those sorts of effects

1386
00:59:01,480 --> 00:59:05,235
can occur directly as opposed
to coming in through the sides.

1387
00:59:05,235 --> 00:59:06,160
Another question.

1388
00:59:06,160 --> 00:59:09,450
AUDIENCE: So they presumably
test some small fraction

1389
00:59:09,450 --> 00:59:10,085
of the modules.

1390
00:59:10,085 --> 00:59:10,710
PROFESSOR: Yes.

1391
00:59:10,710 --> 00:59:16,155
AUDIENCE: That they [INAUDIBLE].

1392
00:59:16,155 --> 00:59:16,780
PROFESSOR: Yep.

1393
00:59:16,780 --> 00:59:19,300
So they presumably test some
small fraction of the modules.

1394
00:59:19,300 --> 00:59:21,120
And then, if those
modules pass the test,

1395
00:59:21,120 --> 00:59:23,290
which take several
months to conduct--

1396
00:59:23,290 --> 00:59:27,080
and that's why there's a large
barrier for module innovation--

1397
00:59:27,080 --> 00:59:28,880
but if they pass
those tests, then they

1398
00:59:28,880 --> 00:59:30,900
assume that every module-- yes.

1399
00:59:30,900 --> 00:59:35,380
So there are a series
of, I would say,

1400
00:59:35,380 --> 00:59:37,790
definitions of what
constitutes a major alteration

1401
00:59:37,790 --> 00:59:39,040
to the process.

1402
00:59:39,040 --> 00:59:41,240
So if you majorly
alter your process,

1403
00:59:41,240 --> 00:59:44,390
then you have to get
your module recertified.

1404
00:59:44,390 --> 00:59:46,500
But if you stay
within those bounds,

1405
00:59:46,500 --> 00:59:50,860
then you can continue selling
under the old certification.

1406
00:59:50,860 --> 00:59:53,280
It's not too
dissimilar to the way

1407
00:59:53,280 --> 00:59:55,170
our food is tested by the FDA.

1408
00:59:55,170 --> 00:59:57,080
It's not every single hamburger.

1409
00:59:57,080 --> 00:59:59,490
It's a spot test.

1410
00:59:59,490 --> 01:00:03,414
In this particular case, these
are examples of the newest

1411
01:00:03,414 --> 01:00:04,830
technologies to
come off the line,

1412
01:00:04,830 --> 01:00:06,586
and you get a couple
of chances to pass.

1413
01:00:06,586 --> 01:00:08,960
If you fail the first one,
it's not the end of the world.

1414
01:00:08,960 --> 01:00:11,490
It just means that you
have to pass the next time

1415
01:00:11,490 --> 01:00:14,160
or else you have a
stain on your record.

1416
01:00:14,160 --> 01:00:17,190
And many of the companies
now have their own facilities

1417
01:00:17,190 --> 01:00:18,569
to do these tests in-house.

1418
01:00:18,569 --> 01:00:20,610
They still send them out
to the certified testing

1419
01:00:20,610 --> 01:00:23,670
bodies that are independent to
get the final seal of approval.

1420
01:00:23,670 --> 01:00:26,540
But they know fairly well
before they send them out,

1421
01:00:26,540 --> 01:00:27,940
are we going to pass or not?

1422
01:00:27,940 --> 01:00:29,498
In the early days
it was-- you always

1423
01:00:29,498 --> 01:00:31,164
had to hold on to the
seat of your pants

1424
01:00:31,164 --> 01:00:32,060
because you never
quite knew what

1425
01:00:32,060 --> 01:00:34,680
was going to happen when
the test results came back.

1426
01:00:34,680 --> 01:00:39,020
But now it's much better,
much more predictable.

1427
01:00:39,020 --> 01:00:41,530
And my favorite test out
of all the ones in there--

1428
01:00:41,530 --> 01:00:45,310
there are a number, if you
imagine a dozen to 20 tests,

1429
01:00:45,310 --> 01:00:48,670
from illuminating it and
measuring the power output.

1430
01:00:48,670 --> 01:00:51,790
One of my favorite tests
involves a hail gun.

1431
01:00:51,790 --> 01:00:56,290
So in a similar manner to
how the FAA tests windshields

1432
01:00:56,290 --> 01:01:00,530
of planes and plane engines
by shooting frozen chickens

1433
01:01:00,530 --> 01:01:02,960
or turkeys at them, in
this particular case,

1434
01:01:02,960 --> 01:01:06,050
we're shooting frozen balls
of water at the modules

1435
01:01:06,050 --> 01:01:08,580
and seeing if they crack.

1436
01:01:08,580 --> 01:01:09,920
That's one of the ones.

1437
01:01:09,920 --> 01:01:12,120
You can envision, as
well, an immersion test

1438
01:01:12,120 --> 01:01:15,850
underwater combined with
bending to see if there's

1439
01:01:15,850 --> 01:01:17,320
any failure in the module.

1440
01:01:17,320 --> 01:01:20,210
So they are tested
quite extensively using

1441
01:01:20,210 --> 01:01:20,900
these protocols.

1442
01:01:20,900 --> 01:01:24,310
And the objective is to
ensure quality and safety

1443
01:01:24,310 --> 01:01:24,990
to the consumer.

1444
01:01:24,990 --> 01:01:26,667
AUDIENCE: What's
the top right test?

1445
01:01:26,667 --> 01:01:27,750
PROFESSOR: This top right?

1446
01:01:27,750 --> 01:01:28,749
Oh, top right over here?

1447
01:01:28,749 --> 01:01:32,310
This is a mechanical loading
test, what it appears to be,

1448
01:01:32,310 --> 01:01:37,452
where you just exert
what is typically--

1449
01:01:37,452 --> 01:01:39,660
that's what it appears to
be in this particular case.

1450
01:01:39,660 --> 01:01:40,368
I could be wrong.

1451
01:01:40,368 --> 01:01:43,770
But my understanding is that
this module is constrained

1452
01:01:43,770 --> 01:01:46,400
on the edges right now and
that there's a load being

1453
01:01:46,400 --> 01:01:48,080
pushed down in the center.

1454
01:01:48,080 --> 01:01:50,670
You could also envision
a sandbag test where

1455
01:01:50,670 --> 01:01:53,780
you load physical sandbags.

1456
01:01:53,780 --> 01:01:58,450
There are a variety of
different types and variants.

1457
01:02:01,250 --> 01:02:01,750
Yep.

1458
01:02:01,750 --> 01:02:02,250
OK.

1459
01:02:02,250 --> 01:02:05,950
So that comes from the
First Solar spec sheet.

1460
01:02:05,950 --> 01:02:08,890
Describe the differences between
various types of PV systems.

1461
01:02:08,890 --> 01:02:10,890
Since we're running
short on time,

1462
01:02:10,890 --> 01:02:13,880
this is a bit of
a repeat from what

1463
01:02:13,880 --> 01:02:15,920
we talked about the very
first day in class--

1464
01:02:15,920 --> 01:02:18,800
the non-tracking, the tracking.

1465
01:02:18,800 --> 01:02:23,110
So I'm going to fly through that
into grid-tied and stand-alone.

1466
01:02:23,110 --> 01:02:24,841
So grid-tied and
stand-alone systems

1467
01:02:24,841 --> 01:02:25,840
is very straightforward.

1468
01:02:25,840 --> 01:02:27,940
If you have a system
that is tied to the grid,

1469
01:02:27,940 --> 01:02:30,245
somehow it has to be able
to interface with the grid.

1470
01:02:30,245 --> 01:02:33,510
Let me go to that in a second.

1471
01:02:33,510 --> 01:02:34,205
Here.

1472
01:02:34,205 --> 01:02:35,830
So if you're having
a grid-tied system,

1473
01:02:35,830 --> 01:02:39,330
that means that your house
or solar array, whatever

1474
01:02:39,330 --> 01:02:41,299
it is that's producing
the electricity,

1475
01:02:41,299 --> 01:02:43,340
has to be tied to the grid
and interface with it.

1476
01:02:43,340 --> 01:02:47,100
Since the electricity
is alternating current,

1477
01:02:47,100 --> 01:02:48,990
there has to be a
device to convert

1478
01:02:48,990 --> 01:02:51,540
the direct current power
into alternating current.

1479
01:02:51,540 --> 01:02:53,020
And that's called the inverter.

1480
01:02:53,020 --> 01:02:55,980
We've already touched upon that
a few times over today's class.

1481
01:02:55,980 --> 01:02:58,960
And these objects range in size.

1482
01:02:58,960 --> 01:03:00,460
For the ones in the
house, you could

1483
01:03:00,460 --> 01:03:02,960
find inverters that
are about yea big that

1484
01:03:02,960 --> 01:03:04,557
can handle a few kilowatts.

1485
01:03:04,557 --> 01:03:06,640
The ones that are handling
these large solar field

1486
01:03:06,640 --> 01:03:09,295
installations, or
segments thereof--

1487
01:03:09,295 --> 01:03:11,170
let me find a large
solar field installation.

1488
01:03:11,170 --> 01:03:11,714
Here we go.

1489
01:03:11,714 --> 01:03:14,130
Here are some pictures of large
solar field installations.

1490
01:03:14,130 --> 01:03:15,264
This is a road.

1491
01:03:15,264 --> 01:03:16,680
These little green
dots are trees.

1492
01:03:16,680 --> 01:03:19,620
That's a big field
installation right there.

1493
01:03:19,620 --> 01:03:24,810
Those inverters are
warehouse-- can be much larger

1494
01:03:24,810 --> 01:03:30,270
sized than the smaller ones
that you'd find in your house.

1495
01:03:30,270 --> 01:03:33,520
And so, listing the major
balance of system components,

1496
01:03:33,520 --> 01:03:38,085
the PV array would be connected
to the grid by an inverter.

1497
01:03:38,085 --> 01:03:40,210
The excess power would be
sent to the grid, causing

1498
01:03:40,210 --> 01:03:41,660
the meters to spin
backward, which

1499
01:03:41,660 --> 01:03:45,180
is a very gratifying
observation from a homeowner's

1500
01:03:45,180 --> 01:03:47,981
point of view or, at the
very least, a rate payer,

1501
01:03:47,981 --> 01:03:49,230
utility payer's point of view.

1502
01:03:52,220 --> 01:03:54,450
The act of sending the
power back to the grid

1503
01:03:54,450 --> 01:03:59,170
is a bit of free riding because
I'm not paying for the grid.

1504
01:03:59,170 --> 01:04:01,654
I'm a consumer, but I'm getting
credit for the electricity

1505
01:04:01,654 --> 01:04:02,820
that I put back in the grid.

1506
01:04:02,820 --> 01:04:05,485
But I'm not helping to
maintain the grid, necessarily.

1507
01:04:05,485 --> 01:04:07,610
I'm paying some transmission
and distribution costs

1508
01:04:07,610 --> 01:04:11,610
on my utility bill, but that's
essentially the same cost

1509
01:04:11,610 --> 01:04:13,700
that everyone else is paying.

1510
01:04:13,700 --> 01:04:15,920
Utility companies have begun
getting wise about that

1511
01:04:15,920 --> 01:04:18,180
and have begun
leveraging surcharges

1512
01:04:18,180 --> 01:04:20,270
based on the input
of solar electricity

1513
01:04:20,270 --> 01:04:21,665
into the grid, which I now pay.

1514
01:04:21,665 --> 01:04:22,990
So I'm no longer a free rider.

1515
01:04:22,990 --> 01:04:25,740
I was for a little bit.

1516
01:04:25,740 --> 01:04:28,240
So, mounted onto the
roof or the ground.

1517
01:04:28,240 --> 01:04:29,800
If you mount the
solar panels, you

1518
01:04:29,800 --> 01:04:31,380
need racking and
framing materials.

1519
01:04:31,380 --> 01:04:34,920
That's also balance of systems.

1520
01:04:34,920 --> 01:04:35,440
OK.

1521
01:04:35,440 --> 01:04:36,570
So.

1522
01:04:36,570 --> 01:04:37,070
Check.

1523
01:04:37,070 --> 01:04:37,280
Check.

1524
01:04:37,280 --> 01:04:37,780
Check.

1525
01:04:37,780 --> 01:04:38,430
Check.

1526
01:04:38,430 --> 01:04:42,340
You typically have, as
well, a circuit breaker,

1527
01:04:42,340 --> 01:04:46,120
or just a very simple lever
that disconnects the solar panel

1528
01:04:46,120 --> 01:04:47,200
from the grid.

1529
01:04:47,200 --> 01:04:49,360
Why is that important?

1530
01:04:49,360 --> 01:04:53,270
The utility companies know
that, on my house in Cambridge,

1531
01:04:53,270 --> 01:04:55,695
there's a solar array
and that that array

1532
01:04:55,695 --> 01:04:57,130
is connected to the grid.

1533
01:04:57,130 --> 01:04:59,700
Let's imagine a tree
falls on the power lines

1534
01:04:59,700 --> 01:05:01,850
during the next storm,
and the utility crew

1535
01:05:01,850 --> 01:05:04,190
has to go out there to
repair it during the day.

1536
01:05:04,190 --> 01:05:06,780
They know that, although the
rest of the grid is down,

1537
01:05:06,780 --> 01:05:09,770
there's that one little power
center on my house that's

1538
01:05:09,770 --> 01:05:11,117
injecting power into the grid.

1539
01:05:11,117 --> 01:05:12,700
And so they'll come
along to my house,

1540
01:05:12,700 --> 01:05:14,641
disconnect the solar
panels from the grid.

1541
01:05:14,641 --> 01:05:16,890
They'll go to my neighbors,
disconnect theirs as well.

1542
01:05:16,890 --> 01:05:18,848
Go to the person across
the street who also has

1543
01:05:18,848 --> 01:05:20,210
solar panels, disconnect them.

1544
01:05:20,210 --> 01:05:23,880
Then go to work on
the utility line.

1545
01:05:23,880 --> 01:05:27,540
So that's why the
external circuit

1546
01:05:27,540 --> 01:05:29,490
breaker is very important.

1547
01:05:29,490 --> 01:05:31,650
Back to inverters.

1548
01:05:31,650 --> 01:05:34,090
Modern efficiencies of
the inverter component

1549
01:05:34,090 --> 01:05:37,970
range between 94%
and 97% typical.

1550
01:05:37,970 --> 01:05:42,770
Record inverter efficiencies up
to 99% have been demonstrated.

1551
01:05:42,770 --> 01:05:45,590
And they all have what's
called maximum power point

1552
01:05:45,590 --> 01:05:49,800
tracking, which means that
they will be able to sense what

1553
01:05:49,800 --> 01:05:52,900
the combined output is
of the array on your roof

1554
01:05:52,900 --> 01:05:56,950
and adjust the
resistance in the circuit

1555
01:05:56,950 --> 01:05:59,070
to match the
maximum power point.

1556
01:05:59,070 --> 01:06:02,070
And that way you're producing
the maximum amount of power

1557
01:06:02,070 --> 01:06:05,150
that you could be, potentially,
under any given illumination

1558
01:06:05,150 --> 01:06:06,050
condition.

1559
01:06:06,050 --> 01:06:08,480
So a lot of fancy electronics
goes into the inverter.

1560
01:06:08,480 --> 01:06:13,360
I really have to give them a
lot of respect in that regard.

1561
01:06:13,360 --> 01:06:15,900
Typically five to seven, maybe
even 10-year manufacturer

1562
01:06:15,900 --> 01:06:16,400
warranty.

1563
01:06:19,040 --> 01:06:21,800
And the mounting
methods-- we talked

1564
01:06:21,800 --> 01:06:24,150
about mounting and racking.

1565
01:06:24,150 --> 01:06:27,230
This type was fairly common
in the early days of solar

1566
01:06:27,230 --> 01:06:29,730
where you'd have these field
installations, especially built

1567
01:06:29,730 --> 01:06:31,780
on hills in California.

1568
01:06:31,780 --> 01:06:34,370
I would say most common now is
really just mounting directly

1569
01:06:34,370 --> 01:06:36,450
on a roof-- oftentimes, flat.

1570
01:06:36,450 --> 01:06:38,900
For large field installations,
one-axis trackers

1571
01:06:38,900 --> 01:06:40,490
are very common nowadays.

1572
01:06:40,490 --> 01:06:42,610
Tracking as a function
of time of day

1573
01:06:42,610 --> 01:06:45,280
and then manually
adjusting the entire array

1574
01:06:45,280 --> 01:06:47,680
for season, if need be.

1575
01:06:47,680 --> 01:06:50,060
So the grid-tied systems
causing meters to spin backward.

1576
01:06:50,060 --> 01:06:52,020
Again, very gratifying.

1577
01:06:52,020 --> 01:06:55,970
And this is an example
of my utility bill, where

1578
01:06:55,970 --> 01:07:00,300
you have a balance
of $163 negative,

1579
01:07:00,300 --> 01:07:03,170
so they're giving me credit
over the course of a year.

1580
01:07:03,170 --> 01:07:07,010
In the winter time-- this was an
example of a month in December

1581
01:07:07,010 --> 01:07:09,930
when a lot of snow
happened and the panels

1582
01:07:09,930 --> 01:07:12,280
were covered for quite a
while, or at least partially

1583
01:07:12,280 --> 01:07:15,480
shaded by the snow for a while.

1584
01:07:15,480 --> 01:07:19,590
And we drew power from the
grid on net that month.

1585
01:07:19,590 --> 01:07:22,920
But on all the other months
we, on net, produced power.

1586
01:07:22,920 --> 01:07:25,590
And negative numbers aren't
shown in the scale here.

1587
01:07:25,590 --> 01:07:28,410
But if they were shown,
you'd see probably

1588
01:07:28,410 --> 01:07:30,830
greater negative value
in the summertime when

1589
01:07:30,830 --> 01:07:34,470
others were producing a lot of
power in the summer-- more sun.

1590
01:07:34,470 --> 01:07:37,250
And at the end of the year,
they draw a line at the bottom.

1591
01:07:37,250 --> 01:07:38,920
And they say, did
you owe us money?

1592
01:07:38,920 --> 01:07:39,420
No?

1593
01:07:39,420 --> 01:07:40,580
All right, great.

1594
01:07:40,580 --> 01:07:43,900
If we owed you
money-- tough luck.

1595
01:07:43,900 --> 01:07:46,670
If we're zero, then
everybody's happy.

1596
01:07:46,670 --> 01:07:48,710
So we typically produce
more electricity

1597
01:07:48,710 --> 01:07:50,640
than we consume in our house.

1598
01:07:50,640 --> 01:07:54,970
And that's an example of what
happens on a practical level.

1599
01:07:54,970 --> 01:07:57,245
Off-grid grid PV systems,
just ever so briefly.

1600
01:08:00,080 --> 01:08:04,224
So let's imagine that you're
running some illicit business,

1601
01:08:04,224 --> 01:08:05,390
and you want to be off-grid.

1602
01:08:05,390 --> 01:08:10,480
Or let's imagine that-- instead,
a more realistic example,

1603
01:08:10,480 --> 01:08:13,870
let's say that you're
afraid of the next power

1604
01:08:13,870 --> 01:08:17,490
outage, of being out of power
for the next 7 to 10 days

1605
01:08:17,490 --> 01:08:20,180
after the storm passes
through your house.

1606
01:08:20,180 --> 01:08:24,790
And you want to isolate yourself
from the grid, if need be.

1607
01:08:24,790 --> 01:08:27,689
You would want to have
a charge controller that

1608
01:08:27,689 --> 01:08:31,600
would be able to distribute
the power to the DC loads,

1609
01:08:31,600 --> 01:08:34,970
or to batteries to store
them, and then the inverter,

1610
01:08:34,970 --> 01:08:37,939
going through the AC loads
or to the grid thereafter.

1611
01:08:37,939 --> 01:08:41,870
So it is possible to architect a
system that is semi-autonomous.

1612
01:08:41,870 --> 01:08:46,350
But it does require more
components and more cost.

1613
01:08:46,350 --> 01:08:49,010
And I just wanted to emphasize
one thing as we close up.

1614
01:08:49,010 --> 01:08:51,310
Life is really good as
an installer right now.

1615
01:08:51,310 --> 01:08:53,899
If we flashed back to
four or five years ago,

1616
01:08:53,899 --> 01:08:57,560
life would have been really
good as a solar panel producer,

1617
01:08:57,560 --> 01:09:00,250
especially if you were involved
in the feedstock industry.

1618
01:09:00,250 --> 01:09:02,450
But now the prices
of modules have

1619
01:09:02,450 --> 01:09:06,979
been severely squeezed
by low-cost competition

1620
01:09:06,979 --> 01:09:08,359
from overseas.

1621
01:09:08,359 --> 01:09:13,729
And the pricing of modules is
around $1.03 per watt peak.

1622
01:09:13,729 --> 01:09:17,700
This is some of
lowest price modules

1623
01:09:17,700 --> 01:09:19,340
that are out there right now.

1624
01:09:19,340 --> 01:09:22,359
And the price of installing
the system on your roof here

1625
01:09:22,359 --> 01:09:24,790
in the US is around $5.20.

1626
01:09:24,790 --> 01:09:27,260
Who's pocketing the $4.20?

1627
01:09:27,260 --> 01:09:30,080
Who gets to keep that money?

1628
01:09:30,080 --> 01:09:31,990
Is it the installer or
the module producer?

1629
01:09:31,990 --> 01:09:33,120
AUDIENCE: The installer.

1630
01:09:33,120 --> 01:09:34,529
PROFESSOR: It's got
to be the installer.

1631
01:09:34,529 --> 01:09:35,080
Right?

1632
01:09:35,080 --> 01:09:38,359
And who's suffering right
now in the industry?

1633
01:09:38,359 --> 01:09:41,090
No matter where you
are-- US, China, Germany.

1634
01:09:41,090 --> 01:09:43,210
Who's suffering right now?

1635
01:09:43,210 --> 01:09:44,350
The module producers.

1636
01:09:44,350 --> 01:09:46,890
The people manufacturing
this stuff right here.

1637
01:09:46,890 --> 01:09:47,870
All right.

1638
01:09:47,870 --> 01:09:50,689
Which brings me to my
current topic of the day.

1639
01:09:50,689 --> 01:09:52,740
Everybody picked up
this one right here,

1640
01:09:52,740 --> 01:09:54,890
this Greentech Media article?

1641
01:09:54,890 --> 01:09:56,190
This is really current.

1642
01:09:56,190 --> 01:09:58,200
November 8, that
was two days ago.

1643
01:09:58,200 --> 01:09:58,700
Eric Wesoff.

1644
01:09:58,700 --> 01:10:02,430
He's higher up there
in Greentech Media.

1645
01:10:02,430 --> 01:10:05,270
And he's writing about
the China-US trade

1646
01:10:05,270 --> 01:10:10,720
war that's evolving pitting--
within the US-- pitting

1647
01:10:10,720 --> 01:10:15,890
the module manufacturers
against the installers.

1648
01:10:15,890 --> 01:10:18,140
So the installers,
these guys who

1649
01:10:18,140 --> 01:10:20,060
are pocketing a lot
of money right now,

1650
01:10:20,060 --> 01:10:21,830
are saying no, no, no, no.

1651
01:10:21,830 --> 01:10:23,250
Let the Chinese modules come in.

1652
01:10:23,250 --> 01:10:24,270
It's great.

1653
01:10:24,270 --> 01:10:26,076
It keeps our costs low.

1654
01:10:26,076 --> 01:10:28,700
The folks who are manufacturing
the modules, on the other hand,

1655
01:10:28,700 --> 01:10:30,560
are saying, well, wait a second.

1656
01:10:30,560 --> 01:10:33,414
They got unfair treatment
by their government.

1657
01:10:33,414 --> 01:10:35,330
And the Chinese module
manufacturers, in turn,

1658
01:10:35,330 --> 01:10:37,660
are saying, well, it's not only
us who are getting subsidies

1659
01:10:37,660 --> 01:10:38,100
from the government.

1660
01:10:38,100 --> 01:10:38,891
Look at yourselves.

1661
01:10:38,891 --> 01:10:39,850
You got subsidies, too.

1662
01:10:39,850 --> 01:10:41,430
And US manufacturers
say, well, you

1663
01:10:41,430 --> 01:10:43,480
got more subsidies than we did.

1664
01:10:43,480 --> 01:10:47,859
So it's evolving into a little
bit of a complex situation.

1665
01:10:47,859 --> 01:10:50,150
There are some people trying
to cut through all of that

1666
01:10:50,150 --> 01:10:55,030
and say, with very simplistic
messages, free trade.

1667
01:10:55,030 --> 01:10:58,100
Or certain other people saying,
made in USA is important.

1668
01:10:58,100 --> 01:11:00,220
Or other people
saying, this isn't

1669
01:11:00,220 --> 01:11:02,375
fair for one reason or another.

1670
01:11:02,375 --> 01:11:04,000
But it's important
to really understand

1671
01:11:04,000 --> 01:11:06,780
the complexity of what's
going on in the context

1672
01:11:06,780 --> 01:11:09,100
of a dynamically
changing market.

1673
01:11:09,100 --> 01:11:12,120
The market is
changing quite a lot.

1674
01:11:12,120 --> 01:11:15,820
In the past, the module prices--
and I say past, recent past,

1675
01:11:15,820 --> 01:11:19,060
as in a year ago-- module
prices were upwards of $2.50,

1676
01:11:19,060 --> 01:11:21,130
$3.00 per watt peak.

1677
01:11:21,130 --> 01:11:23,940
It's only in the oversupply
condition created within

1678
01:11:23,940 --> 01:11:26,910
the last year and a half where
the module prices have dropped

1679
01:11:26,910 --> 01:11:28,850
precipitously down to $1.00.

1680
01:11:28,850 --> 01:11:32,080
And that happens when you're
in an oversupply condition,

1681
01:11:32,080 --> 01:11:37,100
whereas these
installation prices really

1682
01:11:37,100 --> 01:11:38,337
haven't budged too much.

1683
01:11:38,337 --> 01:11:40,920
They're enjoying the fact that
it's in an oversupply condition

1684
01:11:40,920 --> 01:11:43,260
and there's still a
market for the modules.

1685
01:11:43,260 --> 01:11:45,920
So it's complex.

1686
01:11:45,920 --> 01:11:49,970
We'll explore some
of the complex issues

1687
01:11:49,970 --> 01:11:53,630
in our next few classes, and
we'll dive into that in detail.

1688
01:11:53,630 --> 01:11:56,730
I'll leave, for you to study on
your own, life cycle analysis.

1689
01:11:56,730 --> 01:11:58,480
It's very important.

1690
01:11:58,480 --> 01:12:00,330
Life cycle analysis--
in a nutshell,

1691
01:12:00,330 --> 01:12:04,060
it really depends on what you
consider in your analysis,

1692
01:12:04,060 --> 01:12:05,770
where you draw the box.

1693
01:12:05,770 --> 01:12:09,540
And at the end of the
day, module recycling

1694
01:12:09,540 --> 01:12:12,750
is going to be important
as well for determining

1695
01:12:12,750 --> 01:12:17,220
the ultimate impact, both in
terms of CO2 impact, energy

1696
01:12:17,220 --> 01:12:20,647
payback, and environmental
impact of PV in the long run.

1697
01:12:20,647 --> 01:12:22,230
So I'll leave you
with those thoughts.

1698
01:12:22,230 --> 01:12:25,330
And we'll see us on Tuesday.