Semicon 2010

This year’s semiconductor industry gathering, Semicon 2010 is over.  And it was a good show with a lot of technical content targeted at the ongoing effort to achieve ever higher density parts.  The forecast for 2010 and 2011 is for the highest growth levels in a decade.  Certainly, at $295 Billion in projected sales for calendar year 2010, the semiconductor industry is the largest economic activity in the world. And it is just as certainly a more significant economic activity in the US economy than the automotive industry.

Which is saying a lot.

Some of that economic activity is the obvious stuff.  Jobs.  Making things that are important to the industry.  Like all the silicon ingot, water treatment, chip encapsulation compounds, chemical solvents, and gases that are needed.  And all of those feedstocks require people in their respective industries.

There is also the capital equipment market.  Companies that make machines that make chips.  Machines that grow silicon ingots, machines that slice silicon into thin wafers.  Polishing machines that make the surface smooth enough to create the nanometer sized features that become semiconductors.  Wafer probing machines that do functional testing, dicing machines that slice the wafer into the single chips, wire bonding the bare die into lead frames to we can attach the circuits.  Encapsulation, labeling, testing and packaging the final products.

The Semiconductor Industry Machinery business is estimated to be an $11B activity separate from the sale of chips.  The semiconductor equipment market is still the largest target market for motion control products and mechatronics of any market I know of.  At a close second place would be the electronic assembly machinery market  with it’s pick and place, adhesive dispensers and inspection machinery.

Interestingly, the semiconductor industry also provides trickle down technology.  Hard disk drive spindle motors require the exact same 3 phase brushless drive and control as industrial servo motors.  The difference is that the spindle motor is manufactured in quantities of tens of millions of units.  This allows disk drive manufacturers to explore the ultimate boundaries of cost reducing the technology and introducing new techniques to improve performance.  Much of this technology has migrated to the motion control industry in the way of integrated motor control chips.

The semiconductor industry is now made up of two major markets.  Chips and Solar Cells. The solar cell market is counted separately and does not overlap with traditional semiconductor business.  Many of the companies that make semiconductor machinery have extended their capabilities to the solar industry as a way of diversifying into new markets and making up the lost ground that was experienced in the machinery business.

While Solar is still an emerging industry to some extent, it will continue to drive large segments of the economy. Solar photovoltaics and solar hot water drive a lot of jobs in manufacturing and installation of systems.

What we need in the public policy sector is better understanding of the business needs that these industries require.  Generating enough electricity for these industries to thrive is one requirement.  And most states in the US have failed to bring any new capacity on line over the last 30 years. States that recognize these needs and are willing to meet them are going to be the States that prosper with low unemployment and thriving economies.  And that’s where we all want to be.

Intersolar 2010

Intersolar 2010 is one of the larger gatherings of the Solar Energy industry.  I had the opportunity to attend InterSolar 2008 when it was still relatively new.  In the last two years the Solar Energy industry has grown very quickly, chalking up 35% growth in 2009 over 2008 and with similar forecast growth for 2010.  Overall revenue generated from the sale and installation of solar energy systems in the US was estimated at over $2.4 billion.  This is made up of solar panel sales across the residential, commercial and utility customer projects with a mix of technology including some solar hot water systems, some large scale solar concentrator systems and a whole lot of solar panels being installed with racking, inverters, tracking systems, engineering design, contractor labor, etc.

The contrast between InterSolar 2010 and 2008 was very noticeable.  This year’s trade show reflected the growth and sophistication.  Tremendous effort is being put into every aspect of the photovoltaic technology, tandem junction semiconductors that produce the photovoltaic effect at 2 different light frequencies, enhanced surface texturing of the surface glass to improve transmission of light and reduce the problem of incident angle of light, new chemistries like Copper-Indium based photovoltaics which are now “printable” as ink coatings, and ongoing development of thin film silicon, concentrated light focusing on silicon, recycling of the silicon itself, and a host of improvements all targeted at reducing the cost of producing electricity from sunlight.  As you might expect, incredible support from semiconductor equipment makers to provide new equipment to make the technology scalable in production and cost effective in the marketplace.  Solar is the new growth engine in semiconductor equipment.

Solar Panels, like other types of semiconductors, are subject to decreasing cost with increasing volume.  In typical Semiconductor Industry fashion, a lot of capacity has been ramping up since early 2000 which led to a 40% correction (read “drop”) in the price of solar panels during 2009 which played havoc with project bids and created serious difficulties for distributors with inventories or contracts for solar panels at the higher prices of early 2009.  However, now that prices are lower, more demand is expected, and hopefully, companies that had a tough time during 2008-2009 will find business conditions in 2010 and 2011 more favorable.

There is continued optimism that the US solar market will continue to grow at 30%+ per year for the next couple of years, some forecasts are 50% per year and one forecast from Europe suggests 100% growth in the US market next year.  Huge growth forecasts combined with caution on the manufacturing side has created shortages and long lead times for new deliveries.

But we should note, with caution, that this market is largely subsidized by State Renewable Energy Portfolios mandating the alternative energy systems, Utility Company Rebates and Federal Tax Incentives.  Solar energy technology is generally sold on the basis of avoiding future increases in energy costs, not on the basis of eliminating energy costs.

So there is still a big gap in bringing electricity costs down using solar power.  Modern coal plants produce electricity that the utility company can sell at a profit for 6 cents per kilowatt/hour.  By comparison, a 300 watt solar panel will cost at least $1500 installed and functioning.  It can only produce about 756 kilowatt hours per year, and even at 12 cents per kilowatt/hour, won’t break even for about 16 years without government incentives.  So you’re not eliminating your electric bill, you’re prepaying it with a bank loan for a bunch of equipment that doesn’t burn coal.

That’s OK if you can afford it, just don’t make the mistake of thinking you are getting rid of your electric bill.   But be on the lookout for the next breakthroughs in solar.  They are coming.

$2 Bil more for Solar

The President announced $2 billion dollars will be given to fund solar projects in Colorado, Arizona and Indiana yesterday in a radio address.  The funding will pay for several large solar plants that will add permanent power capacity to the respective states.

One report indicated that the $2 billion would be funded as part of the scheduled $863 billion stimulus fund already appropriate by congress.  Another report indicated that the funds would be provided as loans.  There is a huge difference between the two, and the fact that the various reports are not clear on this point is very curious.

As a sidebar, I guess this is the new style of legislation.  The government passes a law first and decides what it means later.

$400 million is provided as a loan (or loan guarantee) to assist Abound Solar to add 2 major manufacturing facilities and new product lines for the company.  One facility in Indiana will be built from an existing automotive plant that will be re-tooled for solar manufacturing.  The company is expected to add several hundred new production jobs.

Abengoa Solar of Spain, which has operations in the United States will be receiving $1.45 billion, although it is not clear if the money is a loan or a grant.  And while Abengoa has operations in the US and has an excellent reputation as a contractor of large energy projects, it seems very peculiar to be giving money to a foreign entity.

This leads to a couple of really important questions about American energy policy.

From the standpoint of cost effectiveness, if you take the $1.45 billion for Abengoa and divide it by the 1500 projected jobs, the cash cost of each job is over $966,000,  per position.   It would be the same as paying $96,000 to each employee for 10 years.  This has to be the most ineffective use of public funds imaginable.

The other public policy question which has come up before is, why are US taxpayer funds being given to foreign companies?  Major green energy projects in every sector are being built by foreign companies with US government funding.   There needs to be a “Buy American” clause in all this pork barrel spending.  If these are loans, or loan guarantees, how does the government get paid back?

The corollary question for US Energy policy is why should the Federal Government be making loans or guarantees to private companies?

Fiskar Automotive, for example, has secured $500 million in loan guarantees from the DOE for it’s electric car program.  But Tesla Motors raised $2.1 billion in the private financial markets.  Does this constitute a scenario where the Federal Government is creating unfair competitive conditions by providing financial support to companies of their choosing?  And not providing similar funding to other companies.

This is also true on the larger scale.  As the Federal government continues to direct where the majority of US research and development funds will be spent, the process itself disconnects the efforts of the research community from the potential economic benefit that the research should be targeting.

The goal of all research is to produce a benefit.  And the benefit must be weighed in the context of economic utility.  When the development of technology is subjected to bureaucratic decision making, it is dissociated from the decision making process of economic benefit.

This will result in massive waste as limited resources are put into projects with poor return in value.  We appear to have entered a period of time where the process of free market decisions are being circumvented, and everything is to be decided by government.

Because, after all, these folks are professionals at spending your money and they know better than you, or the market, what is most important.

Solar Tracking

Like all things mechatronic, solar tracking is hard to describe.  If we consider the actual motion, it’s two degrees of freedom and both motions are rotary.  The problem is to rotate a flat rectangular panel both about it’s midline, which is azimuth tracking, and rotate it about its baseline, which is elevation.

And the two motions are essentially simultaneous.  Yes, we only see the azimuth motion because it is the daily motion of the panel.  But the angle of elevation has to be mechanically available at the same time so that the annual change of the sun’s angle to the earth can be adjusted.  You could probably get by with this one by going out and mechanically adjusting it 4 times a year, and it would work fine.  It’s just an extra hassle, and if you’re going to bother doing tracking it might as well be good tracking.

Here’s the reason tracking is so important.  The National Renewable Energy Lab says that dual axis tracking can add up to 36% to the energy harvest of photovoltaic panels.  That’s a big number.  It’s a bigger number than anything that is in the lab dealing with the fundamental efficiency of the energy conversion process.

Because tracking the sun has such a big impact on energy harvest, it gets attention.  There are about 20 companies and tracking systems around the world.  There are all kinds of interesting solutions to the mechanical problem.  There ought to be a prize for the best design.  Some of them are really wild.  But all of them have one thing in common.  They all move arrays of panels instead of one panel at a time.

There are two main areas of solar tracking, concentrating solar and photovoltaic panels.  Concentrating solar systems are generally arrays of mirrors that focus the sun’s energy on a target area to produce high temperatures that generate steam and turn a generator.  Large arrays of mirrors all pointed at the same spot require constant adjustment and very high precision in order to get the sun’s energy concentrated on the right spot.

In photovoltaic systems, the panels convert light to electricity directly and need to be perpendicular to the sun.  But accuracy of  +/- 1 degree is acceptable.  So in one sense, it’s not as difficult.  But 2 axes of rotation is a difficult motion problem to solve.  So there are a lot of solutions out there.

I spent some time working on this and there are several really simple solutions that are possible.  And in the process of researching all the possible solutions, we found a wide range of mechanical systems that are available.

Solid Tech Inc. is in the process of developing a cost effective solution that does 2 axis solar track on a single solar panel.  This approach serves commercial flat roof installations and residential applications increasing the total energy harvest and reducing the payback period for the system including the cost of tracking.

Stay tuned for more details.

Solar Power and Economies of Scale

If solar power costs decline to “grid parity” or the same cost as generated electricity costs at the grid, it will take over a significant portion of the utility industry.  That has been the goal for 20 years.  It’s a great idea.  Because eventually homeowners can generate their own electricity, become independent of the utility and reduce their operating costs.

Or so the story goes.

We’ve been trying to reach grid parity for some time.  Without much success.  And not because we aren’t trying.  Billions of dollars of government subsidies, R&D funding and private investment are being poured into the pursuit.

Energy independence!  Both as a Nation and as individuals.  It would be great to be able to say, personally, we don’t have pay any utility bills.

The first great fallacy is that either solar power and wind power can cause energy independence for the US.  This is because the energy we depend on is not electricity, it is Middle East Oil for gasoline.  We are dependent because of our cars and the choice to not make our own gasoline, even though we can at lower cost than importing it.

But on top of that, almost none of the electricity generated in the US uses Oil.  It’s all coal, natural gas, or nuclear.  So the idea that the US will reduce it’s foreign oil imports by generating electricity with solar power or wind power, is completely ridiculous.  There is no connection between the two.

There is a theoretical energy equivalency that can be expressed.  But there is no real connection.  So people who make this claim are intentionally misleading anyone who listens.

How are we doing with respect to the cost of electricity generated by solar power?  It’s been an interesting couple of years.  The industry experienced a brief shortage of raw silicon which kept prices fairly high.  More recently there was a precipitous drop in panel prices.

Opinions vary as to the cause of this drop, but with the massive increase in manufacturing capacity worldwide, I would guess that the price drop is strictly a matter of oversupply.

Economies of Scale will fix the problem according to some.  After all, look how well we’ve done with computers, hard drives and flat screens. Flat screens that were $10,000 to $50,000 a decade ago are now affordable to the point where the CRT has become obsolete.

Since the biggest component cost of the solar panel is silicon wafer, we should expect similar results in the solar market.  The stampede to build more solar panel manufacturing plants resulted in oversupply.

Now the race continues to drive costs down.  Panels that were selling for $3.50/Watt a year ago are down to $2./Watt and prices are expected to continue to fall. And some manufacturers will not be able to keep up with falling prices using older technology.

But are we getting to grid parity?  Is Solar power cheap enough to compete with utility power?  Nope. Because even at today’s bargain pricing, a 225 Watt panel will only produce 900 kilowatt hours in a year at maximum efficiency.  At market cost for electricity, $.05/kW, it’s only $45 worth a year.  And it currently costs about $1125 pay for the panel, installation and balance of system components.  That means it will be around 25 years, the end of the useful life of the system, before it breaks even.  Yes, in California where consumers pay $.23/kW the payback is better, but it’s still very expensive to convert to solar.

We have got to do better than that.  And we will.  The technology is coming along.  But economies of scale by themselves can’t quite get us there.

Scientists Create First-Ever Circuit Powered By Light

For the first time, scientists have created a circuit that can power itself, as long as it’s left in a beam of sunshine. Created by scientists from the University of Pennsylvania, the world’s first photovoltaic circuit could eventually power a new line of consumer devices or even model the human brain.

Right now the creators can only coax minuscule amounts of electricity from their photovoltaic circuits, far too little to power consumer electrical devices, although those amounts could quickly skyrocket.  There are plenty of other ways they say that they can squeeze more electricity from light. Right now only about 10 percent of the photovoltaic circuits on a glass side work. Increasing that number will boost the power output.  Another way to get more power is by turning their 2D structures into 3D structures. Stacking multiple layers of light-collecting and electricity-using circuits would also boost power.

The photovoltaic circuit is a scientific breakthrough, not a technological one. These new circuits will most likely never replace their silicon counterparts.

Photovoltaic circuits could be ideal for other applications, however, such as powering tiny robotic devices or running computer calculations at the speed of light.  Far into the future, these circuits could even be used to set up as artificial neural networks that could model the brain.

At their most basic, computers represent data as on or off, a “0″ or a “1.” Using light instead of electrons, these photovoltaic circuits could store data from, say, one, two, three or four. Each number would correspond to a certain wavelength or color of light — red, green, blue and yellow, for example. To model the human nervous system, each color of light could correspond to a different neurotransmitter, say red for dopamine and blue for serotonin.

The potential applications of the technology are huge, but will take years to develop into any kind of practical equipment.

www.news.discovery.com

The Road to Recovery

Economic recovery, job creation, whatever you want to call it, everyone is trying to figure out how to get employment numbers back up and get the economy back on track.  Seems that the situation is pretty severe, and we’ve managed to export it to all of our trading partners.  But there is controversy over the information being reported.  The news is in the position of impacting the tone, so they can make things sound bad, or not, depending on who’s data they use and how they phrase their comments.

Many analysts have commented that the reporting of employment data in the US has been manipulated over the last few years and, for example, there is a whole category of people who are unemployed and are not being counted because they are assumed to have quit looking for work.  (How’s that for cynicism)   And China has been accused of artificially holding it’s currency value low in order to minimize the impact of decreased exports.   Interestingly, US exports are up significantly due to the weak dollar, so there’s at least some silver lining to the clouds of an economic downturn.

But the real road to recovery is based on real value.  Technology is great, it enables a lot of new product concepts that make our lives more convenient.  But the root is in the value that is delivered.  That value can be something compelling, like the iPod, which offers the convenience of incredible portability and simplicity in delivering entertainment media, or a major improvement in the energy storage capacity of batteries which makes hybrid cars possible.  The value that a product delivers is what makes it attractive and drives a customer to own it.

This means that technology cannot be dissociated from it’s economics.  The electric car is still a challenge because a cup (about 20 cents worth) of gasoline contains enough energy to move a 2 ton car down the road to the gas station when you’ve run out.   Which is a very cost effective exchange instead of you and several of your friends pushing your SUV to the next station to fill up.  So an electric car should cost about the same to operate as a gasoline car unless there are many people who are willing and able to pay a premium to drive electric.  And it’s getting there, and there are quite people who will pay the premium.  About 300,000 a year now.

But there is no product on the market that is immune to improvement.  So the thing that will really get the economy moving, and get people working, is product development.  Any improvement that enhances the value of a product is important to the success of the US as a nation.  That means everything.

That’s why the current dialog in the alternative energy sector is so important.  Significant improvements such as direct drive generators in the wind market and lower cost tracking technology in solar offer big advantages in the overall economic performance of these technologies.  This will reduce the amount of tax dollars that have to be used to subsidize the emerging industries.  And I’m betting there are a lot more improvements to come.

So the road to recovery is improving value.  In any product, in any market.  And that’s change we can really count on.

Defining Green Jobs

The American Solar Energy Society (ASEA) reported a staggering forecast of “37 million jobs resulting from Renewable Energy and Energy Efficiency in the US by 2030″.  That’s pretty exciting stuff.  In fact, the numbers are so big, I had to get a closer look.  37 Million new jobs would certainly fix things in this, or any, economy.

I know from having lived in Colorado that there is a lot going on there because there are 300 sunny days a year, so it’s a great place for solar power projects.  Many businesses like the Whole Foods company and other large retailers have projects going on and the Federal Center is putting in major solar power arrays to reduce their electrical demand .  The National Renewable Energy Labs are in Golden Colorado, near Denver, and they have huge campuses with thousands of people working on all sorts of energy related topics.  Public Service, the local power company has major wind farm projects going on in the state.   And there are major wind testing facilities that have recently sprung up to support the wind industry in Colorado.

So as part of measuring the impact of Renerwable Energy and Energy Efficiency activities, the report from the American Solar Energy Society focuses quite a bit on the local Colorado situation.  I have heard that Colorado claims to have 10,000 new jobs in the renewable energy arena.  There is a long way to go to get from 10,000 jobs to 37 million.  More importantly, the report states that  5600 of those jobs are at NREL and other government or non-profit organization, like ASEA.  I don’t think you can count jobs that are paid for with taxes, they don’t produce revenue.

The methodology of the report includes under it’s definition of Renewable Energy category of business; wind, photovoltaic, solar thermal, hydroelectric, geothermal, biomass (ethanol, biodiesel, biomass power), fuel cells and hydrogen.   That’s pretty broad by itself.

Then the Energy Efficiency businesses include; appliance, HVAC, insulation, automobile and other.  So on a statistical basis some of the appliance and HVAC businesses, which already exist, get counted for a certain amount of head count and revenue generated because portions of their product sales focus on energy efficiency.  Wow!  That means you can count a small percentage of almost everything else.  Personally, I don’t consider guys putting insulation into your home part of the emerging green economy.  But the folks at ASES do.

And jobs that are federal, state or local government that are related to renewable energy are counted as well.  Non-profits, trade associations, foundations, consultants, investment, and other related positions are all in the count if they are related to Renewable Energy or Energy Efficiency.

So there is a huge gap between estimating the “impact” of the green economy and job creation based on the green economy.  Looking at the report from ASEA, they are clearly not the same thing.  If 56% of measured jobs in Revewable Energy in Colorado are in the government and non-manufacturing roles, then the claims for job creation can not be more that 4400.  And the size of the “green” job creation opportunity is significantly lower than the announcement would lead you to believe.

Worse still for the US economy, one of the dominant suppliers in wind power is Siemens, which means that a lot of the sales generated here, translates to revenue for Europe.  In the solar cell market Japan, Germany and China supply the majority of the market and several foreign suppliers have operations in the US.

We sure hear a lot about how the green economy is going save our economy.  I think some of the messages are exaggerated and taken out of their proper context.  The analysis needs to be based in terms of what we can really expect in Growth.

Response to Big Wind

There were a number of emails relating to the posting on Big Wind a few weeks ago and I would like to make some clarifications.  My numbers were off, primarily based on the fact that “retail” rates for electricity such as you and I pay, are much higher than the wholesale rate which utilities pay.  The actual cost of electricity is around 4 to 5 cents depending on a variety of conditions that can influence the rate.  Based on that fact, the estimated payback for a large scale wind turbine is 2.5 times longer than my estimate in the posting.

In an effort to check the facts, I visited the American Wind Energy Association website.  There is a brief article (www.awea.org/pubs/factsheets/EconomicsOfWind-Feb2005.pdf)  that goes into some detail on the costs of large wind projects.  I used some of that information for a reality check.  They put a utility scale project analysis in the article that starts with a 50 MW capacity.  They say that a project this size would probably cost around $65 million to install.  I put in the recently published national average $1.93 million dollars per megawatt and got $96 million, a whole lot more expensive.

If the turbines are located on a site with wind available 35% of the time, and that’s a big number, the site should generate 150 million kilowatt hours per year.  The revenue will be around $6 to $6.75 million dollars a year not including maintenance, property taxes, management costs, etc.  Best case, the project will break even in about 10 or 11 years.  Worst case, 15 or more. Don’t know if the life expectancy of the equipment is 20 or 25 years.  I’ll check into that.

And by the way, medium size photovoltaic projects operate on about the same basis.  They are too expensive to make it worthwhile to take your home off the grid.  So the DOE’s “Million Roofs on Solar” project is doomed.  You can legislate a program, but you can’t make it work unless it pays for itself.  And the government doesn’t have enough money to pay for it for you.

These are really big numbers to be tossing around, and it makes me uncomfortable that the mainstream press doesn’t report on the facts.  Maybe that’s too much to ask.  It is a somewhat technical subject.

OK.  Here’s the real point.  The projects have many financial incentives like State and Federal subsidies, accelerated depreciation and investment tax credits helping to subsidize the costs.  That’s how we get to payback periods of 6 to 8 years.    So my math was wrong, but the final results are as represented.

When politicians make policy without facts, it comes out wrong.  Wind is great.  I am an avid wind energy enthusiast.  But we cannot create these massive alternative energy industries with billions of dollars of commerce that require government subsidies in order to operate.

I have done some numerical analysis of the wind energy problem and there are solutions that will result in 2 year paybacks for investors.  We need more innovation that integrates financial responsibility in the equation.  This is the kind of engineering that is self sustaining and will result in new jobs without government subsidies.

Big Wind and the Absolute Cost of Technology

The American Wind Energy Association published results for last year’s spending on wind energy.  The US spent $16.4B on new wind tubines and installed 8500 megawatts of generating capacity.  That’s $1.93 million dollars per megawatt of capacity.  That’s a lot of money.  Especially when a megawatt of capacity of wind energy may only produce 300kW of actual power based on the amount of wind that can be harvested.

The efficiency rating of a wind generator is not related to the equipment, but rather to the average wind speed and number of hours out of a year’s time that the system is generating power.  So this number can vary quite a bit, and of course, the generated electricity varies with the wind.  So a lot of effort is put into the site survey to determine if a particular location can generate enough power to pay back the cost of the equipment.

At 30% efficiency the average power generated is 300kW.  This is enough electricity to power 231 homes if the homes are all using about 1300 kWh per month.  Personally, I have not been able to get my power usage under 200kWh per month, so it might be many less homes in actual practice, but you get the idea.  If you are paying 11.5 cents per kilowatt hour that’s only $149 per month in electricity.  So the revenue for 1 megawatt of capacity is $34,535 per month.  And since a wind farm has operating costs, usually estimated at 10%, the revenue minus operating cost is $31,082.  To pay off that $1.93 million invested will take 62 months.  Sure, it will go a lot quicker if the electricity rate is high like in California.  But it looks like everyone is making money at this alternative energy stuff except the consumer.

Texas has very low energy costs to begin with, and solar power has slightly lower net efficiency than wind power due to the number of hours of daylight, the number of days of sunshine, etc.  So the local utility has begun suggesting to customers that because of expensive investments in wind and solar alternative energy systems, that we (the customers) will have to pay increasing rates for power to “help shoulder the costs”.   Really?  I thought all this alternative energy stuff was going to lower our costs.

I’m not a financial genius, but I can tell there’s a problem.  Especially when no one in the alternative energy industry ever talks about return on investment. We have to focus on technology that has better financial performance.  And I think it’s out there, and my company is working on some of the solutions.  We’re just stuck behind the slow moving giants of the industry who are dominating the landscape.  It’s time for some of that Yankee Ingenuity to come to the forefront.

Next Page »