Shining Light on EMCORE

| About: EMCORE Corporation (EMKR)
This article is now exclusive for PRO subscribers.

Let me ask you a question: When you think of Boeing (NYSE:BA), General Electric (NYSE:GE), Intel (NASDAQ:INTC) and NASA, as an American how do you feel? These three companies have provided Americans with the means for long distance travel for business and to see loved ones, great jobs, financial security, innovative computing, electrical and health products; in fact, you could credit these companies with making great contributions to improving your quality of life. And in the case of NASA, inspiration for all Americans and pride in their country at what their countrymen can achieve.

Now what if I told you they may have been working towards something in common, so exciting that the way the world generates electricity from the sun need not be the same ever again?

I’d like to introduce a new name to you: Emcore (NASDAQ:EMKR) Corporation of Albequerque, New Mexico. What is it that is so special about the great American organizations above and Emcore, and what do they have in common?

The answer – CPV or Concentrated PhotoVoltaics solar systems.

What is CPV?

CPV, or concentrated photovoltaics, is a new terrestrial way to harness the suns power and convert it in to an energy form we use every day – electricity. Standard flat plate solar photovoltaic panels have been around for a long time you may think, but great leaps forward in PV in mass production, cost, size or roll out have not occurred.

CPV promises to change all that for three reasons:

1. Concentration up to 2300 suns

A normal photovoltaic panel consists of a flat plate of silicon (or other similar products), usually under a flat bed of glass. The area of silicon to produce 1 watt of power is dependent on the sun on the day, the quality of the silicon used, direct or non-direct exposure to the sun and of course the technology used and its respective conversion efficiency. Now to give a brief example of where PV technology is currently at, I will use First Solar's (NASDAQ:FSLR) last annual report as a guide. 457,000 panels to produce 30MW of power equates to 65 watts per panel. Now remember this uses all panels at a cost of around $4.81 US and $4.71 Euro for the plain module. First Solar’s efficiency is currently at 10.6%. Now this is not meant to mislead you as others, such as Sunpower (NASDAQ:SPWR), obtain efficiencies up to 23.4%.

But as First Solar is the company which many people do know about and they have one of the highest current installed annual manufacturing capacities at 308MW, they have been used as an example. Remember this 308MW, as I will come back to it.

Most of the traditional terrestrial (on earth) means of photovoltaic electricity production from the sun rely on a 1-1 ratio. That is the glass used is simply a means to enclose and protect the silicon (or Cadmium Telluride) cells underneath from the elements.

I’d like you to think about a magnifying glass or more importantly a lighthouse lens. The goal in these instances is to manipulate light so that it is able to be concentrated on a single point or smaller area. In the case of concentrated photovoltaics, there are two ways this is done. The first is by utilizing what is called a Fresnel Lens. In the case of a lighthouse, a circular, complicated lens has been used to achieve this. But what if you could manufacture a lens from flat glass which achieves a similar objective of concentrating light? Well Emcore, with its development partners, has achieved this with concentrations upwards of 1000 suns. For an idea of what I am referring to, take a look at the small diagram on Sol3g’s site showing a flat piece of glass and how it redirects light to a very small area.

The second way is to use a reflective arrangement, but that is not currently how Emcore’s integrators intend to concentrate the sunlight, so it is irrelevant for this posting.

2. Triple junction cells

The second innovation Emcore has been working on is adopting their space application solar cell technology for use back here on Earth. Current standard photovoltaic technology is designed to harness a single band of the suns energy. If you look at the spectrum of light generated by the sun, you will see that the predominant bands are the blue, green and red spectrums.

Emcore's photovoltaic technology differs from flat panel in that instead of capturing and utilizing only a single band of the spectrum, it actually captures 3 bands - the Red, Green and Blue - instantly increasing the energy available for use by around 300%. Emcore is a market leader in the development and production of these triple-junction cells and according to their conference call on 8 May 2008, currently has 9 MOCVD reactors installed and two production lines operational producing 20MW of these cells per month. Its current best efficiency levels of its cells are approximately 39%, however it is continuing development on the technology with a potential efficiency substantially higher than this figure.

Now you may now be thinking that the production of these cells would be more expensive than standard silicon, and this would be a correct assumption. But then if you consider that a) these cells are designed to operate with over 1000 times the energy input into them with the concentration technology, and b) they are 39% efficient, the numbers start to look a little better. For interest sake lets look at 1cm2 of this type of cell. Raw material cost is lower than standard PV as instead of using silicon as the base, it uses Gallium Arsenide or GaAs.

Gallium is extracted primarily from the residue obtained during the processing of aluminum and secondly via electrolytic zinc. Other sources include fly ash collected from burning coal.

Now consider the ratios. Instead of one sun's concentration, we have 1000 suns' concentration. Instead of 10.6% efficiency, we have 39% efficiency. (I’ll round down to 3 ½ times). So for the same 1cm2 area of a CPV cell versus a standard flat plate cell, we get 3500 times the power generated. Even using Sunpower as a base efficiency of 22%, we still get 1500 times the watts from the same area of cell. Now if you think this can't be right, please take a look at the diagram comparing PV flat plate vs the CPV cell size on Emcore’s site.

In other words you would need to produce 3500cm2 or 1500cm2 the amount of panel to compete with 1cm2 of Emcore cell. So on a strictly capital base for machinery and sq ft area you need to produce the same output of MW off the production line vs standard PV, this in itself will make available substantial capital cost savings. The figures which I can provide are for a combined 1GW production line utilizing Emcore and Suncube/Sunryder technology the cost is approximately $120-$150M USD. For an equivalent sized flat panel production line it could be as much as $1200 M USD.

3. Tracking the sun

The third innovation concentrated photovoltaics has introduced is a two-axis sun tracking system. This ensures the photovoltaic array is fully exposed to direct sunlight from the moment the sun rises to the time when it sets. The tracking systems Emcore and its development partners use have been field tested and refined over the last two years to a point where power usage, breakdown and manufacturing costs have been optimized.

CPV Consortium

Now having explained the technology as well as I can above, you might wonder how Emcore is planning on introducing this new technology to the market. Well first, a little history on the development of a number of crucial affiliations with other industry players who are also advancing their development of CPV.

ISFOC contract for 3MW of generation over 2 years was awarded to these companies:

Isofotón of Spain (700kW), Fraunhofer ISE spin-off Concentrix Solar of Germany (500kW) and SolFocus (500kW), incorporating multi-junction III-V-based PV cells made by Emcore, Azur Solar and Spectrolab...

..... and four more firms to install the second phase of a further 1.3MW of CPV systems: Emcore Corp of Albuquerque, NM, USA (300kW), Arima Eco of Taiwan (300kW), and Spain’s Concentración Solar La Mancha S.L. (300kW) and Sol3g S.L. of Cerdanyola (400kW).

Following the CPV summit at ISFOC in Madrid, the CPV consortium was officially declared. This consortium has one very unique interesting common goal.

No technology exclusivity

How this is implemented may in fact provide some clue to why it seems Spectrolab (Boeing) and Emcore, which would normally be considered competitors, have demonstrated such close associations.

One of Spectrolab’s main customers, Solfocus has some interesting venture capitalists. NGEN partners include Asahi Glass, Boeing, Henkel, Air Products, BASF, DSM, Schott, Canon, Saint-Gobain, Bayer, Dupont, Siemens, Bekaert, Honda and Unilever. All multinational with a combined total revenue of over $600B a year.

The CPV consortium, with its significant backers, completed prototype arrays, capacity for mass manufacturing and exposure to marketplaces the world over will be one of the driving forces ensuring mass adoption of this new technology.

What stage is the technology at?

Lets take a brief look at Emcore’s 25kW array power curve versus the National Grid (NASDAQ:UK) power curve. Now consider this. Emcore’s array begins generating power around the 7.30am mark, reaching full output at approximately 8am. This is maintained until approximately 4pm with a rapid drop in output until 5.30pm. Now the National grid electricity demand shows a similar early morning demand beginning at 7.30 am, peaking at 8am and continuing throughout the day, the only difference being the period from 4pm until 10pm, after which the power demand substantially reduces. In effect, the stability of this solar technology, could in practice supply on-demand power for 8 of the 14 peak hours demand per day. And there are technologies being developed which allow excess capacity to be stored quite conveniently for use at non-generating times. But that is for another posting.

Now Emcore has not been idle, nor is it intent on producing the completed CPV systems on its own. The model it appears to be following is one very much similar to modern silicon computer chip manufacturers. In that Emcore will in fact manufacture high-end complete CPV systems, but it is also willing to work with external OEMs to develop their product lines which use the Emcore concentrated photovoltaic cell as the heart of their system. Think Intel, IBM (NYSE:IBM), and AMD (NASDAQ:AMD) for similar instances of how this business model has been made to work in mass coverage of the market.

At this stage, Emcore has received orders from OEMs including GGE, Concentration Solar La Mancha, XinAo and ES Systems of South Korea but is keeping relatively quiet about further associations. (More news links). But they would be developing more of these business relationships with OEMs, you can be sure. A list of integrators using this CPV technology can be found here.

Feed-In tariffs are the current key to launch new industry

Now if we also take a look at the current feed-in tariff say in one of the countries currently aiming to get a decent installed base of Solar generating capacity, South Korea where ES Systems are based, they have a guaranteed $0.75 USD per Kilowatt hour feed in tariff. Using the 25kW array above and cutting off the ramp up and ramp down, we get 8 hours x 25(kW) x $0.75. Or $150 per day. Multiply this by 365 and you get $54,570. And how much does one of these units cost. You’d have to contact Emcore, but pay back time, even including installation, maintenance and breakdown would be under 2 years. And as both Emcore’s cells receivers and Korea’s feed in tariff is guarantee for 20 years, you can see the financial incentive for ES systems to get their product lines operational ASAP.

Now you could argue that there is a time limit on the guaranteed feed-in tariff and that it is expensive power in the case of South Korea, however this is a future orientated policy, which in itself, provides immediate investment incentive for the country to entice solar manufacturers into the marketplace.This means new production facilities and jobs for the long term as this CPV technology can only become cheaper. I will explain my reasoning behind this bold statement later in this article.

Manufacturing Capacity

In the introduction, I referred you to the current installed MW (Mega Watt) production capacity of First Solar (FSLR), a US $21.65 billion company as being at 308MW current. Now Emcore, during their conference call on 8 April indicated they have two production lines operational with the third to be producing cells by June. And ….

…we have a manufacturing capacity of 600,000 units per month or 20 megawatts per month by the end of this quarter.

This means within two months, the cell production capacity from Emcore will be 240MW per year, however they are building a fourth line in China with undeclared manufacturing capacity to be operational by September of this year. In effect, their cell output within two months will be 78% of the total MW capacity which First Solar is currently producing.

Emcore and General Electric (GE)

In 1999, Jack Welch, then head of GE, founded GELcore a solid state lighting [LED] joint R&D effort with Emcore. This business division was established to develop low cost, innovative, low energy lighting solutions. The joint venture ran for 7 years until Emcore’s 49% share was then completely sold to GE in August 2006 for $100 million.

Some would say the relationship ended there. But instead consider this – GE and Emcore work together for 7 years on a product line. Their Engineering, Management, production staff all work together to develop a new technology. The technology is successful. Emcore then gets $100M to move on to focus their efforts on developing something else. What do they pick? Well, terrestrial CPV of course.

At this point you may wonder why, if CPV is so revolutionary, does GE no longer has a stake in Emcore’s success? Well they do. And even more telling is when they have elected to buy back into Emcore.

On April 15 2008 GE purchased 230,020 shares in Emcore. And on 30 April 2008 tripled their holding to 690,060.

Now this in itself may seem unremarkable, as GE own shares in all sorts of companies, but I would say that you are missing the point. GE had a history in R&D with Emcore supposedly up until August 2006. Less than two years ago, Emcore started developing CPV at the same time. It is logical to conclude that not only would people from both organizations be still associated with each other and what each company is doing, but also by GE recently buying back into Emcore, that they believe in this technologies prospects and in doing so have declared their interest and confidence in Emcore management.

Note: the mutual fund listing needs a second take when you search by date (or activity) and look at all of the investment houses and companies who have recently either added or increased their stake in Emcore. But that is not the point of this article.

Lastly, with regards to whether GE’s recent purchase of Emcore is a strategic decision or not, consider first GE Energy decision in September of 2007 to invest in PrimeStar Solar Inc. Then consider GE’s attendance at the 33rd IEEE Photovoltaic Specialists conference from May 11-16th. Note this one is a little hard to put together, but look at page 7 and who Jim Rand is, then look at the two presentations on page 13, May 13th at 10am in Area 3 and Area 4.

Now reconsider why, before this conference had even begun, GE tripled their stake in Emcore?

Emcore and NASA

Emcores associations with NASA go back a long time. Space was the first frontier for developing solar generated electricity with satellites and space stations relying on solar power to run the equipment where other energy sources could not be used. Even the mars Pathfinder used Emcore (Tecstar) cells to supply the unit with power.

And although Boeing via Spectrolab supplied ATK the cells for the current Mars Phoenix Lander Solar lander, Emcore is working with ATK to develop the next generation Ultraflex-175 solar array.

Emcore’s relationship continues to this day with ongoing contracts with NASA.

Emcore and Boeing (BA)

As discussed above, not only have Boeing via Spectrolab, NASA and the CPV consortium developed this CPV technology in conjunction with Emcore, Emcore has actually supplied Boeing satellite division with their cells. Back in 2004, Boeing contracted Emcore to supply its satellites with their 27% efficient solar cell. Emcore, with Boeing’s CPV cell division, Spectrolab, has also contracted a common wafer supplier in Sylarus.

Also present in Madrid was Sylarus Technologies, a startup that claims to be a source for the germanium substrates used by Emcore and Spectrolab in their cells.

In fact, in a lot of instances the associations, joint technology development, common suppliers and the CPV consortium confirm that the ties between Boeing’s Spectrolab and Emcore run more than just skin deep.

Emcore and Intel (INTC)

Now the Intel connection did not really make a great deal of sense until I saw this article about IBM developing CPV with a concentration of 2300 suns. So a natural thing to consider then is if a chip manufacturer such as IBM is developing this technology, wouldn’t it make sense for other chip manufacturers to be developing this as well. So I then found another article and found these lines.

Both solar and semiconductor industries are similar in that they rely on silicon, integrated circuits and precise manufacturing techniques.

Semiconductor equipment maker Applied Materials Corp. is in the midst of a major expansion into the solar and photovoltaic markets. Intel Corp., the world's biggest semiconductor maker, also is exploring investments in the solar business”

And then I went digging. It seems in August 2007, Intel were reported to be looking for potential solar systems manufacturing technologies.

Here's a thought - what if the Intel association with Emcore is not what it seems?

This is the heart of the CPV system.

Now consider what has happened to the microchip industry with relation to Moore's law, cost and mass production of silicon wafers and Printed Circuit Boards.

What if a similar potential exists here?

What if it is a simple retrofit to change the Intel production lines from producing microchips and mother boards for PC applications, to one which produces GaAs chips and control boards for CPV technology?

The manufacturing capacity, distribution network and associations with OEMs Intel has in place means this could in fact be their reasoning behind the FO division swaps for shares in Emcore.


There is a lot to consider when you are guided by the pundits about Emcore’s business plans, particularly their focus on Emcore’s optical division and the line of questioning which comes from ‘research’ institutions about their client base. But I for one believe that Emcore has it right. They have developed their product to the point that there is certification available with an IEC standard; done the hard yards by building a number of prototype arrays with various integrators with different ideas; have set up a financing and business model which brings the capital outlay to build a full 1GW production line down by a factor of 10; and having built 9 MOCVD reactors and 2 production lines with 2 more already being built, they have the in-house management and engineering expertise to build many more of their product lines. There is no reason to not believe Emcore in their grand plans for the concentrated solar marketplace.

Their backers and consortium partners, and their current and past associations with Emcore, speaks volumes for their credibility, if not the potential for this technology to finally come of its own and provide truly grid parity (or cheaper) electricity generated from nothing more than the sun.

A great sun play of America, if you ask me.

Disclosure: Author has a long position in EMKR