In the two years it has been in India, the American solar
panel manufacturer, First Solar, has gathered as much mind-share as market
share. The company has been in the news for both positive and negative reasons.
On the positive side it is seen as an aggressive company that has quickly
bagged orders from developers and EPC contractors.
Today, it has about a fifth of the market under its belt.
The country’s first grid connected utility scale plant — Moser Baer’s 5 MW
plant in Sivaganga, Tamil Nadu — has First Solar’s thin film panels. Recently,
the company secured an order for 25 MW of panels from Green Infra, a sizeable
order in the solar industry.
On the negative side, First Solar is seen as a company that
has succeeded on the back of some aggressive lending by the US Exim Bank. Besides,
the company’s Cadmium Telluride-based technology is frowned upon by
environmentalists and its order book has been fattened by an ill-advised skew
in the procurement policy of India’s largest solar programme. The National
Solar Mission allows developers to import thin-film based modules, but mandates
local procurement if they opt for crystalline silicon technology.
Sujoy Ghosh took over in May as the country head of First
Solar India, which contributes 8 per cent to First Solar’s revenues ($2.8 b in
2011). Sujoy earlier worked for GE and Tata Honeywell. In an interview to
Business Line, Sujoy speaks on the maths and myths of First Solar. Excerpts:
What is your order book position today?
Today, there are 225 MW of operating assets with First Solar
modules. There are several more under construction by developers who won
projects in Batch-II (of the first phase of the National Solar Mission), but I
would not like to disclose the details because it is up to our customers to
disclose them.
We now have about 20 per cent share of the Indian solar
market and we will maintain the market share.
You have incorporated your company here. What is the idea?
When we entered India in 2010 —that’s the time National
Solar Mission was taking off — we had a fairly satisfactory run of the market,
in terms of selling modules to either third-party developers or EPC contractors
— first to some of the ‘migration projects’ and then to Batch-I of Phase-I of
the NSM and then Gujarat. Over the two years we have developed confidence in the
Indian market that solar is here to stay. Solar has a space in the overall
generation mix of the country.
What prompted us to open an entity is we want to build local
capabilities on the ground and have a much broader relationship with the solar
market in India.
Could you amplify on that?
Sure. When I say local capabilities — look at First Solar’s
capabilities. We have the expertise to develop projects, do EPC, get financed,
both on debt and equity, maintain those assets for the investors. The thing
which we like about the development business is it helps to develop a pipeline.
While the programme in India has been successful, it is lumpy in terms of
demand. All the PPAs get signed on one day, everybody wants COD on the same
day, decisions are delayed till the last moment, either on expectations of a
drop in prices or to work through the documentation. This creates stress on the
system in terms of people like us or EPC contractors who have to stock
inventories and keep people on the bench, or let go of the opportunity.
For us as a company, our manufacturing process is
continuous, unlike crystalline silicon. We are still the lowest cost
manufacturer because we run out production lines optimally to full capacity to
hit those cost points. If we run out lines only three months in a year we
can’t.Therefore, it helps if we start developing our own pipeline
of projects, it helps us to bring predictability of demand which then enables
us to leverage our backend organisation — not just manufacturing, but system
design and EPC and plan for building capacity as it comes. That helps us become
more competitive.
That’s what prompted us to look at India as a key market.
One of the five ‘sustainable solar markets’ for us based on their natural
economic need, good irradiation or the country wanting to conserve their fossil
resources are India, Australia, Saudi Arabia, South Africa and Chile — India
and Chile because of the high penetration of diesel.
Our technology enjoys a good footprint in India and we want
to broad base that relationship and get into a bit of development from our
side, bring in our systems engineering expertise. A lot of these plants — the
first wave of plants, either under NSM or Gujarat — have been built, in a lot
of cases, by EPCs who don’t have much experience in doing solar. Yes, there
have been a lot of European EPCs who have plenty of solar experience in terms
of building assets, but there have also been an equal number of home-grown EPCs
who have come in and built plants.
I think asset quality in some cases might have been
compromised. Fifty per cent of the cost of energy is the funding cost. I think,
as a stakeholder in the industry, it is our duty to make lenders comfortable
with the asset. Only then will they lower their risks and only then will we see
good quality capital. Right now, many of these projects have been built with
recourse to the balance sheet. That is not really project financing. For Indian
banks, ‘will it work’ is the question that they want to see because there is
not much record of generation. Large Indian lenders, for instance PFC and REC,
who lend to other parts of the power sector, have not really got into solar and
unless that happens you cannot sustain an industry in India.
Do you organise funding for your customers?
Our customers have been able to organise finance, a
combination of local banks and foreign banks, we are not privy to the nature of
the …
There is an impression that you have been successful in
India because of the backing you have received in the form of low-cost
financing by the US Exim Bank.Well, first of all, US Exim is not captive to First Solar —
there are other US manufacturers in the market besides First Solar.
The other thing is — while I am not sure about the exact
percentage — a majority of our 225 MW is not financed by US Exim. The number of
projects financed by US Exim is probably a third.
The point is, we are selling because our technology, at
least in hot climatic conditions, produces more power than polycrystalline
silicon.
We hear very divergent views on that.
See, it is like this. Our nameplate efficiency figure is
about 12.7 per cent. The poly guys maybe at 15.5 to 16 per cent. The mono
crystalline may go up to 20 per cent. All these efficiencies are at ‘25 degrees
C, one atmosphere’ conditions. Now, each manufacturer publishes what is called
a temperature co-efficient. The degradation curve — as your ambient temperature
rises, your efficiency will fall. For the First Solar modules, that degradation
is 0.25 per cent per degree rise in temperature. For the average
polycrystalline, it is 0.45.
So, as the ambient starts hitting 40 and above, our modules
start producing more. This is ambient 40. Typically, the cell temperature is
15-20 per cent higher than the module temperature. FS modules typically produce
8-10 per cent higher energy under high ambient conditions.
The second thing about thin films generally is that the
impact of diffused sunlight —cloudy conditions, or dusty conditions — that
causes poly output to drop further, compared with TF. India is a combination of
high ambient and diffused. So generally we find that we get a higher energy
yield in India. That’s been fundamentally the reason for our success. That
helps to lower the LCOE (levelised cost of energy).
Thin Films, in general, and First Solar panels, require more
space. But the incremental cost of land is outweighed by the incremental yield
In how many months in a year in India would the temperature
be higher than 40 degrees?
Rajasthan and Gujarat practically nine months in a year.
First Solar modules are best suited only for these states
then?
In other places, even if the ambient temperature is lower,
the cell temperature is 15 per cent higher. If the ambient temperature is 35,
then the cell temperature is 50. I am saying, at more than 40 degrees our
panels start producing more electricity.
The other impression going around is that thin film has a
market in India because of the skew in Government policies (that permits import
of thin film modules, but requires crystalline silicon modules to be made
here.)
In Gujarat there is no such policy lacuna. The National
Solar Mission was 150 MW, and Gujarat was 600 MW. We got more share in Gujarat
than in NSM competing against Chinese companies.
I think people have bought us primarily because of our yield
performance. Second, they see First Solar as a profitable company. In projects
that are financed on non-recourse basis, the lenders demand a great amount of
due diligence and look at the solvency of the supplier for enforceability of
guarantee and warranty obligations.
True, US Exim has indeed helped people who have bought from
us. But then exim financing is available from other countries too. Exim is not
a captive product of the US. Exim is an enabler, but people first make a
technology choice. If they go for Chinese technology, they have access to the
same level of funding from the Chinese banks. It is unfair to say that the only
reason why we are successful or we got this kind of installed base in India is
because of EXIM.
Some experts say that thin film modules made sense when the
price differential between thin film and crystalline silicon was large. But now
the delta is so small and hence thin film modules do not make sense.
That argument is absolutely correct under temperate
conditions and rooftops. Because you end up paying higher price for ‘balance of
system’ for marginally higher price of the modules. But because of the yield
and because of the diffused sunlight and relatively low cost of land in the
overall economics of the project, there is still a significant amount of
advantage that the thin film brings.
The other point that we must remember is that crystalline
silicon pricing seen in the market today — in our view and as is evident from
the balance sheets of the companies — they are selling below their
manufacturing cost. So, the question is whether this cost is sustainable.
So, you have the balance of systems penalty for TF and the
higher yield advantage. Net-net, thin film is still advantageous.
Within thin film, do you think your Cad-Tel technology will
continue to rule the market?
Within thin film, we have CIGS, Cad-Tel and amorphous
silicon. I don’t know whether you are aware or not, First Solar was also
looking at CIGS, very actively, trying to build an alternative technology. The
cost of CIGS in our view is at a point that it will take them a long time to
catch up with us.
It is one thing to say ‘we have hit a certain efficiency
level in the lab’ and quite another thing to say ‘we are bringing such
efficiencies on a sustainable basis in our production line’. That’s a
considerable gap.
Second, are the efficiency levels (claimed by CIGS
manufacturers) bankable? How much data do you have to back that efficiency? Any
lender will say, you have reached that efficiency in the lab, fine, but show me
where it is working.
Crystalline guys would then have the same argument against
you? Do you have the same performance data as they do?
Why, we have performance data for 16 years.
In India?
In India also we have data. In India, even the crystalline
guys have date only for two years, because the grid-connected plants are only
two years. Off-grid is not a real measure of efficiency.
What about amorphous silicon?
In today’s world any new breakthrough technology requires a
lot of money to scale up commercially. The incumbent technology people will
also continue to improve their technology. Even we are working on our
technology to improve our efficiencies. It will be incremental.
On the poly side, you put in more material you will get more
efficiency. If you scan the global solar scale, you might see a lot of
interesting concepts. But if a company like GE, which was pursuing its own
solar programme gave it up, I’m sure other smaller players will find it very
difficult.
Thin film is really about the manufacturing process. How you
lower your rejection rate? How you are able to consistently produce the same
module-after-module, deposit the same material in the same way so that you get
the same results? It is about consistency in the process. It takes considerable
amount of time before you hit that. Till you hit that there is a lot of
rejection.
We have passed that point. Relatively we are pretty
stabilised.
Is the market then a blue ocean for you?
I am not saying that. It is not a blue ocean. India has got
other challenges. We are just discussing technology. I think from a technology
standpoint we have got certain advantages when it comes to utility scale solar.
That’s one dimension.
The challenge is our competition pricing their products at
probably below cost. I think the bigger challenge, as these programmes get
built in India is about ‘will this policy sustain’? Will the grid keep pace?
You can build a solar plant, but you do face grid congestions.
Finally, solar has to reach a point where wind is today, for
it to be really sustainable. Pricing of power, cost of power from that aspect,
has to be something where you need a FIT (feed-in tariff) to keep the wheels
turning. As an industry, all of us are trying to look into that point. Because
all of us do believe there is a shortage, and if we can bridge some of the gap
— it may not be huge in energy terms — people might be willing to pay a higher
price because right now the alternative is no power.
Do you see a situation where for a want of funds for FiT,
the industry is pushed to only bilateral PPAs?
The National Solar Mission is going to be there. The
Government has made a commitment.
But where is the money for NSM?
Look at it this way. The average price of Batch-II was Rs
8.50. The average HT price is at Rs 6 — and in the southern states even if you
pay the price you get power for three days a week. So, there is a demand.
Question is, are people willing to pay more for the demand? Second is, how much
are they willing to bet on the future price of coal. Solar is a good hedge
against that, because fuel is free.
A lot of us in the industry are trying to discover a demand
outside of FiT. But at the same time I think the NSM will be needed to give
scale. Indian industry needs to scale up. For that, both forms of programmes
(FiT and bilateral PPAs) need to co-exist.
Who is going to pay for the FiT?
Today, the gap between the price of conventional power on
the higher end and the price of solar power coming from FiT is not that big.
The differential has come down. You should intuitively believe that coal prices
should go up. Who predicted coal prices will be $140 a tonne. That’s where you
hit the grid parity point.
The question is, even if grid parity happens, will it uncork
large demand? Just by doing grid parity, it may not. You need other things such
as open access, how you are going to wheel and bank this power … really the
grid and the grid operator being comfortable with this kind of power coming in.
The solar assets which one creates also need to have features that make them
dispatch-able.
We are building assets such as a 550 MW A/C in the US. These
are large chunks of power coming into the grid. There is a well defined grid
code which we have to install in the design of the solar plant which will make
this power-grid-friendly.
I think as the plants in India, NSM Batch II come up, people
will begin to see some of those features being installed. Same like wind. From
the 250 kV class machines today you have 2 MW class machines. A 2 MW machine
has got a lot more features in terms of voltage ride-through, voltage control,
and some of those features which allow the machine to not go off-grid when
there is a grid disturbance — it rides through the disturbance.
A solar plant is also similar in nature. In wind you have
what’s called a converter. In solar it’s an inverter. It’s virtually the same
thing. What these Government programmes will do is, to help people to scale,
demonstrate that the power can be integrated into the grid properly, which will
give confidence.
Bilateral deals will happen, but again, we will not see
bilateral deals such as 100 MW happen in year one. People will test the waters.
Even if I have a PPA — even if I am generating, who is assuring me that the
grid is available? That’s where in this whole ‘bilateral deals’ we will still
have to figure out how to take the grid companies along with us.
So you still continue to depend upon these
Government-sponsored programmes?
We need both. I am saying that while the Government
programmes are important to scale…
Look at it this way. We’ll see probably the first few deals
happening on bilateral basis. If it works, I am sure there is enough demand for
people to move into these bilateral deals.
Today it is not about the appetite to take power — there is
the appetite. There is appetite to put the assets up. The key is really the
grid.
What are First Solar’s ambitions in India?
Our goal is to see if we can create a pipeline which is
predictable enough for us to then figure out if the Indian market is mature
enough for us to put in manufacturing operations here.
How do you intend to create a pipeline?
By developing our own projects. We are in discussions. As
you would recognize, when we do development … development and EPC construction
requires a lot of local expertise. We do believe that in India there is enough
local expertise available already.
So what we are trying to do is to structure partnerships.
Like: you got local developers who want to build projects. First Solar is here.
We’ve got the technology, got the experience to do this. We know how to put a
quality asset together. We also have some leverage in terms of financing. So we
are trying to put our combined expertise together instead of recreating
competencies.
Does it mean that you will not own the assets?
We ideally would not. We are not an IPP. We are a technology
company. What we want to do by development is really try to enable a pipeline.
Are you going to adopt the ‘wind model’, where a
manufacturer puts up a wind farm and then sells slices of it to investors or
IPPs?
We are trying to make that a model. The challenge for us,
unlike say Suzlon building a farm, is the price of power. Also, the Suzlon
model was built around ‘accelerated depreciation’. Whereas what we are trying
to do is to enable more solar energy coming into the grid. Slight differences,
but in the end the model remains same.
We want to create the assets and then ultimately transfer
the ownership to somebody who wants to own it long term and then have an
O&M agreement so that we deliver our performance for as long as the guy
wants to be comfortable with the asset. This is what we do in the US, Canada
and elsewhere.
At what stage is this thinking?
It would be premature for us to disclose to plans, but we
are looking at States where there is good solar irradiation, and States which
have good demand. Bilateral PPAs make more sense in South India. Currently,
that is where the power constraint is more. The Southern grid is starved of
energy. The bilateral PPAs will immediately make sense here.
What are your major concerns?
The big question mark is the enforceability of the RPOs
(renewable purchase obligations). It has not really happened, even if it
happens in the industrial sector —because in energy terms they are a big
consumer — if you start enforcing it across discoms, how are you going to
monitor, there is not enough monitoring capacity available — even if you
enforce it on a limited scale on the industries, I think that will give a
fillip to the market. REC/RPO is a good programme, but without enforcement it
has no meaning.
We are in the initial development stages with some partners,
looking at land options, looking at how we get the optimal PPA pricing and as
soon as we sort out some of these things we could probably begin construction.
Will the removal of accelerated depreciation for wind power
developers work in your favor?
There is some interest from some of those who have AD
appetite, but to my knowledge no deal has got closed. But remember, these AD
customers will start placing orders in February. So, it is early days yet.
What is your take on the manufacturing in India? Should the
Government mandate local procurement or not?
Our view, as a developer, is that there should not be any
restriction on their ability to source stuff. For the manufacturers, it is a
function of predictable demand — solar manufacturing is all about scale. Also
cost of power, cost of utilities in India is extremely high. We have to really
see what kind of incentives Government is giving to people who are putting
capital. That is secondary. The biggest issue is, is there enough scale
domestically created which can justify manufacturing at the same time not
impact the developers.
Because, if you set up a smaller scale manufacturing
plant…globally there is over capacity. The solar developer community is trying
to lower the cost of power so that it creates more demand, and at this point in
time if you impose restrictions on them, saying they have to source only from
domestic, they lose the advantage.
Therefore, it is a function of a) on the front-end, how do
you create scale, which will come if policies are consistently enforced.
Enforce RPO, you immediately create a market. If it is a natural economic need,
definitely people will come and put up manufacturing. We cannot have and should
not have a policy where you restrict the ability of the developer to bring to
you the lowest cost of power. If you do that, if cost of solar power goes up,
it is going to shrink the market, and it will be a non-starter. It will have
exactly the opposite effect of what you are trying to create.
Do you think there is scope to bring down costs outside
modules, say, in Balance of Systems?
I am not an expert on that subject, but here is what I would
say: the amount of cost focus that has been there has not been there on the BoS
front. But at the same time, remember, the BoS has got like fifty different
components.
Take the biggest — the inverter — is there a good scope for
cost reduction?
Even if you take 30 per cent cost of the current
inverters—inverter is like 2 cents per watt in an overall project cost of, say,
$ 1 .75. It is not going to make much difference.
But the module is only 50 per cent of the overall cost. In
some cases it is just about engineering, more than taking cost out — things
such as structures. In the first wave, the good ones, the EPCs mostly came from
abroad, because there was no Indian. They adopted the global designs to do the
first level of projects. People have learnt from that.
The Indian EPCs looked at their experience in non-solar and
built something. They have learnt from that. The way you do electrical systems,
foundations etc. There is scope to take BoS costs out, but a lot of it is about
how you engineer the plant and get more efficiency around engineering—how do
you design the plant for Indian conditions. People have learnt a lot.
What would you say about the high warranty claim reserves
set off by your parent company?
Any semi-conductor material which goes into higher
temperature will experience more abuse. It is physics. What we said was
historically 90-95 production used to go to temperate climates. Over the past
few years, India had about 8 per cent of our global revenues; we are building a
lot of plants in the US in very hot conditions. What we said was as our demand
changes from 90-10, to exactly the opposite, we will have more warranty
reserves, because we could naturally expect some recall of projects, because of
the more abusive conditions. That’s all to it.
