The deal for the 360MW of extra generation capacity, for a total of 960MW, comes right on the heals of a 500MW cancelled solar power development in the Mojave desert due to environmental concerns. This deal is better suited because it is located on private lands, which aren’t subject to the regulation and scrutiny as when companies build on BLM and other public lands.

The Coyote Springs development is located north of Las Vegas, straddling the Clark and Lincoln county lines. Originally imagined as a distant suburb with more affordable housing (at the time) than Las Vegas, the recession, high gas prices and the long commute (60 minutes each way) have put the damper on those plans. By using these private lands for solar power generation, Coyote Springs can secure income for that land, and BrightSource can cut a lot of the red tape out of the process of building solar power facilities.

There are transmission lines nearby, however most of the power would feed into Las Vegas – one of the issues I’ve been reading up on lately was the transmission gap around Las Vegas. There is transmission lines coming in from the north terminating near Apex, and then again from the south to California and the Eldorado Valley (where there are many other solar power projects), but nothing through the city.

[via Reuters]

The phase one portion of the study included narrowing down areas with renewable resources to the zones where the best resources in each state were available. This includes areas of high solar irradiation and sufficient wind speeds at 50 meters above ground. The WECC (Western Energy Coordination Council) peak load was 150GW in 2007, so the resulting 180GW of generation capability identified by the study could significantly mitigate peak load.

Other factors that were included in the study were sufficient density – enough to provide 1,500MW of energy in a 100 mile radius, which would justify a 500kV transmission line from the zone to the demand region. Also excluded were protected areas identified by the BLM and other federal government agencies where the development of renewables is prohibited, not congruent with the development intent or could impact sensitive environmental conditions. Also not included were smaller renewable installations – generally at the local level like PV panels on rooftops.

After all is said and done, the 11 western states identified 80GW of solar thermal, about 74GW of wind, 2.5GW of biomass and 29GW of geothermal, for a total of 185GW (though only about 162GW of that was in WREZ areas).

The 11 western states in the lower 48 in 2007 consumed 667,670 GWh. In the same states, the renewable zones identified could possibly contribute over 442,000, or 66% of the total power consumption. Some states like Montana, Wyoming, New Mexico and Nevada could export renewable energy to states that wouldn’t be able to produce anywhere near what they consume – California.

Future phases of this study will analyse and create a transmission plan, as well as developing a regional purchasing system for utilities to meet their states’ RPS (renewable portfolio standards) goals, and finally getting the various local, state and federal jurisdictions to work together harmoniously in the permitting and construction process to get renewables from generation to distribution.

Solar PV is often looked to as a simple solution – put the panels on your roof and you can generate between 50 and 100% of your power requirements over the year. It simplifies a number of things – like treating the grid like a gigantic battery, where energy generated that isn’t used locally is sent out to the grid and when the panels aren’t producing enough, to pull off the grid to make up the difference. The issue is that if you have a large PV installation designed to generate power, and the sun goes behind a cloud and your output drops by 50%, the power demand doesn’t instantly drop by that much, so you have to make it up elsewhere. This “elsewhere” is usually natural gas, but natural gas can only spin up so fast – from 50% to 90% in 6-7 minutes, by then the sun might be back out from the cloud and production rates are already returned to normal. This limits the amount of large free field PV that can be installed – its not so much about cloudy and rainy days since you can forecast and plan for additional resources to be run that day, its the intermittent cloudy days where your output varies all day. Geographically distributing the generation can help but building large facilities far away from each other can hurt the total cost effectiveness of the project. I would put the overall piece of the generation pie at 2.5% of total energy generation.

The proliferation of roof-top systems could double these numbers if utilities can get a better grip on the demand side – being able to turn off non-essential grid demands like plug-in cars, and being able to ramp power quicker when PV generation quickly drops (draw on energy storage – either batteries, thermal or kinetic). This is where the smart grid is absolutely critical. Also, predictive systems could also cut consumption – your air conditioning unit sees that the sun is behind a cloud so it can reduce energy consumed because your house isn’t exposed to direct sunlight and wont heat up as quickly as it would have otherwise, instead of just turning on when the internal temperature creeps up.

Solar Thermal is an easier approach from the grid operator’s perspective. Since solar thermal is still driven by hot liquids and turbines, it has a higher generational “inertia” than solar PV does, that is to say the output doesn’t change as quickly as PV does when the sun goes behind a cloud. This would allow for other natural gas turbines to speed up and add generation output. Likewise, its possible to build storage in the form of a large molten salt tank that could be drawn on when the sun goes behind a cloud, as well as after the sun sets. A 100MW facility that had 50MWh thermal storage could provide for up to 30 minutes without any sun, or a proportionally longer time during partial sun exposure. Thermal storage is still far preferable than batteries for the foreseeable future. It wont be until there are large quantities of partially used batteries to be recycled for other uses (old PHEV batteries that are retired after 10 years) that battery storage will become more useful. This technology has a substantially larger piece of the pie – 12.5% of total energy generation.

Wind has an upper limited of around 20% with today’s “dumb” grid. With the creation of a smart grid and devices that can defer their energy consumption to cheaper times or whenever there is extra capacity, wind will be able to grow beyond the 20% barrier. Beyond that, other large scale storage mechanisms like pumped hydro would allow for wind energy to be converted into kinetic energy in the form of water stored in a reservoir, however this is a fairly expensive exercise – instead of just the cost of the wind turbine and transmission lines, the additional pump storage system costs have to be factored into the total cost of wind power. Wind could account for about 20% of energy generated.

Geothermal technologies aren’t really limited by reliability, since they substitute for baseload power (coal, nuclear). Rather its limited by the land suitable for geothermal wells, mostly in the western parts of the US. Standard geothermal requires pumping hot water (at least 300° F) from underground and letting it turn into steam to turn turbines and then either releasing the steam into the atmosphere or attempting to recapture it and put it back into the ground. Even lower temperature geothermal, which can use water down to about 165°F, still is limited in areas it can be used, though it provides a larger footprint in terms of how much energy we can extract. Enhanced Geothermal involves fining hot rock and then artificially recharging the aquifer with water to be able to extract that hot water elsewhere and use it for generation purposes. Combined, these technologies would only account for about 5-7% of installed rated power, but because of their high capacity factor provide around 15% of our energy generation portfolio.

Biomass, using waste from humans (trash) as well as marginal plant products (the famous example is  switchgrass) to produce power is also an option, however it is further behind from a technological standpoint that solar or wind. However it could function as a baseload power plant, which is highly desirable from a CO2 reduction standpoint since it can replace coal power. The biomass is gassified and then put through a turbine to generate electricity. I don’t see this ramping up soon, the projects are currently very small and no where the size for what a utility would require for a generation facility (100MW or larger).

Combined, all of these green technologies can account for approximately 50% of our energy generation needs. When combined with nuclear and hydro, we could have 80% of our energy from non-fossil fuel sources. However its not going to be that easy to shut off coal – only geothermal and biomass have sufficiently high enough capacity factors to replace coal. Combined, they’re likely to only be able to displace about 30% of coal’s energy output.

To eliminate coal will require a very large, geographically diverse wind power footprint – with a larger capacity than I think the current grid can support. To replace 70% of coal with wind power, it would take approximately 450GW of installed capacity – higher than most estimates have predicted.

Any final drive to remove the last few coal plants by 2030 will require replacing them with baseload natural gas power plants, which are subject to the volatile prices of natural gas (in 2008, natural gas was $9/MMBTU, now its approximately $3.50/MMBTU, I have no idea where it’ll be next year).

We should always work for a cleaner environment, but ignore the reality of technology at this point is not advisable. Where alternatives exist (geothermal in the west, solar in the southwest, wind in the midwest) we should shun coal power, but we don’t have the option to turn it off completely. Cap and Trade could push us further in that direction, but it shouldn’t have the end goal of eliminating coal, rather seeking to reduce CO2 – which would have the effect of preventing coal power expansion and forcing coal burning facilities to do what they can (including building newer, comparatively cleaner) to clean up.

Even with FirstSolar recently announcing production costs below $1/W, there are still many other companies that haven’t yet managed to hit that price point.

[B]y the end of 2009, year-end cell capacity will stand at a staggering 17.6 gigawatts. When taking into account production and capacity ramp rates, this is enough to produce almost 14 gigawatts of modules. By contrast, we estimate demand for 2009 to be only 5.5 gigawatts.

The analysis breaks up the solar industry into three groups: those that make it, those that wont, and the ones in the middle.

SunPower vs First Solar 2009 (Google Finance)

Vertically integrated companies and thin-film CdTe panels make up the first group, and they’re pretty much assured of making it through. How well they get through is a different story – looking at the stock price for SunPower (SPWRA) and FirstSolar (FSLR) you can see that SunPower is down 20% and FirstSolar is up over 40%.

Other companies, those using exotic and inefficient $/W designs are in line to fail unless they can have the cash to survive some very rough quarters.

Finally, the group in the middle is those with promising but not yet proven technologies that are working to lower $/W. They are headed towards either buyouts if they have decent technology worth acquiring, or closing up shop.

The only real flaw I find with the analysis is that the ranking of companies is based on the $/W module price. Some types of panels have higher installation costs per watt, and whether or not the system is one-axis tracking, annual kWh/panel, etc. All of these factor into the final ROI. But it is still the most comprehensive analysis I’ve seen.

The problem can be distilled simply: if you have 1MW of PV generation, and the sun goes behind a cloud, how quickly does that 1MW go down, and how far does it go down? My experience with solar PV shows that the time is nearly instantaneous – when the panel is shaded, the output drops immediately. The how far down part depends on the type of cloud. If its low, dark storm clouds, the panels go from nominal output to below the inverter’s threshold of power generation instantly and no power will be generated.

Think of this from the utility’s side – they have 500MW of load to match with 400MW from a natural gas generation facility that can output up to 600MW, and 100MW from solar PV panels that are next to the natural gas plant. If the panels get shaded over the course of a minute, the natural gas plant cant just turn up from 400MW to 500MW immediately. This is due to the nature of the combined cycle generation – modern units have spin-up rates around 7% per minute. So 7% of 600MW is 42MW, so it will take a little over 2 minutes to get back to 500MW of generation. By then the cloud could be gone.

Even if you had software to control the turbines to match the per-second output of the PV array, if the array managed to get shaded faster than the rate the combined cycle system can spin up additional power, then you’re going to have issues matching the load with the power generation.

There is one key to this situation though – how fast can the panels get shaded. If you have a cloud moving overhead at 100 feet per second, and the cloud is 1000 feet long, thats roughly 10 seconds of shade and a 1000′ long shadow on the ground (note I’ve made some assumptions on a few things, like the sun is pretty much stationary over a 10 second period). There are a few things that follow from this…

• Does the turbine get more wear and tear from spinning up at its fastest ramp rate for 10 seconds, the going back down?
• How much PV output gets shaded? Even at 10,000′ long, depending on cloud shape and plant layout, it might only be shading a portion of the total panels, leaving other panels producing at nominal output.
• Does a larger geographic distribution of panels provide for better coverage against the occasional cloud?

There are many questions that need answering when it comes to distributed PV generation. One idea I’ve had is look-ahead demand shaving. In other words, you’d have to build some sort of system to look at the sun and see if there was a cloud coming soon from the perspective of the PV plants. Then, based on the color of the cloud, get an idea how much output will fall, and then send that back to the grid operators. A running net total of PV increases and decreases for the system would be calculated every few seconds so the grid operators would be able to know if the PV-based power output is about to go down, or spike up after the sun comes back out from behind the clouds. They could either use this data to unlock additional power (say, from a flywheel or battery storage system) or send out a message to grid users to try and shave their demand for the next 15 seconds (or whatever time frame is calculated). The backend calculations could get very complex – knowing what data is coming from what geographical position and create a virtual cloud map of the skies above the city, and even doing its own predictions for when dips and spikes will happen.

The problem boils down to if you can geographically distribute the generation, what kind of reliability issues are there on intermittent cloudy days, and how much of that power generation drop can you predict ahead of time to schedule additional generation resources.

I should note wind faces a similar problem. I was looking at a study for how 15GW of wind power in Texas would affect the ERCOT grid. It turns out the maxium 30-minute decrease was around 2.8GW (93MW/minute), which would happen once every 3-4 years, and 2.4GW about once per year. ERCOT, which has a peak demand of about 65-70GW. To accomidate the 2.8GW 30-minute decrease at a 7% ramp rate for combined cycle, you would need to have that much in spinning reserve because even fast starting plants will take about 30 minutes to start. If a plant is at 40% (minimum operational capacity) and had to spin up to 100%, at 7%/min thats less than 10 minutes. So its completely possible to have sufficient reserve within a dispatch pattern, given enough individual turbines (approximately 5GW in the previous example). Also, the report notes that this condition is limited to certain seasonal patterns, so its likely that the dispatch configuration needed would only be required in certain situations.

Abengoa currently has a 11MW solar power tower, and now they’ve moved up in scale to construct and operate a 20MW tower. The tower has 1,255 mirrors focusing the sun’s rays onto the tower’s boiler to convert the water into steam to turn turbines and generate electricity. Abengoa has plans for up to 300MW of solar thermal by 2013.

The next step in the development of concentrating solar power isn’t just bigger and better towers, its also the use of thermal storage. Heating molten salt in tanks can store that thermal energy until after the sun has sent and traditional solar power is no longer available. The way that the electricity demand curves look (at least on the west coast of the United States) during the spring, winter and fall is a peak during the daytime, as well as a peak between 5-7pm as many businesses are still open, but people are also coming home from work and turning on lights, appliances, etc. This double whammy causes demand to spike after the solar generation peak has passed.

In order to increase the penetration of solar power to more than just the daytime hours, you need to store energy to be recovered later. Gigantic batteries are far too expensive, and pumped hydro requires the necessary geography or artificial structures to have water travel a long vertical distance.

In December of 2008, Sempra was proud to commission their existing 10MW thin-film based power generation station in the Eldorado Valley near Boulder City, NV, near the existing Nevada Solar One 64MW solar thermal facility. The energy from the 10MW is sold to PG&E of northern California.

Now, only four months later they’re announcing that facility’s first expansion up to 58MW, an addition of 48MW and approximately 800,000 First Solar thin-film solar PV panels. The expansion is contingent on a power purchase agreement. Sempra Generation said that it had available cash to fund this project and isn’t relying on financing, rather they will self-finance.

First Solar, the PV module manufacturer, announced in February that it had crossed the $1/W manufacturing threshold. The $1/W price tag comes with a catch – the thin-film panels produced by First Solar are less efficient than traditional PV panels from a land area perspective. While traditional PV manufacturers offer panels that ouput between 225 and 315W, First Solar offers smaller sized panels (0.72 vs. 1.6 sq meters) from 60-77.5W, roughly one half of the solar conversion efficiency.

This results in more cost per W installed when compared to traditional PV – more panels, more freight, more labor to install, etc. However, given the race between traditional PV and thin film, it appears that thin film is leading by a hair in terms of costs.

The report (PDF) is available and goes through the all of the steps on how to calculate the LCOE and what factors go into designing a large scale solar power system. There are a few places where I disagree with their numbers but overall the report is fairly accurate (their maintenance figures a little low – not too bad, but for us our maintenance cost per kWh is not close to one cent or half a cent as they might claim in some of their cases).

There are a few highlights to point out in this report. First was a reference to a report on panel degradation (source report PDF). They tested 23-year old solar panels and found that they had only degraded 4%. Further, there was nearly no noticed degradation from years 1 through 20, with nearly all the degradation coming in a few year window between years 20 and 22, with the last year of the survey having leveled off the degradation.

Next is that most plants are financed under forecast power production, and that is usually grossly underestimated. This I have also found to be true – our guaranteed output is far less than our actual output – by more than 10%.

The biggest item in the report is the following quote…

In SunPower’s case, the grams of polysilicon consumed to manufacture a watt at the solar cell level declined from 13 g/W in 2004 to 6.3 g/W in 2008 and is planned to decline to an estimated 5 g/W with SunPower’s Gen 3 technology now under development. By 2011 this approximately 60 percent reduction in the use of silicon, coupled with an approximately 50 percent decline in the price of polysilicon, will independently drive large cost reductions for PV panels.

So while panels might have cost $5-6/Wp back in 2004, the increase in cell efficency, reduction of the quantity of bulk silicon used as well as the reduction of the cost of silicon due to the crappy economy and oversupply due to added manufacturing capacity, the cost of a panel could drop down to $2/Wp, and reducing overall costs from $7-8/Wp to $5/Wp and closer to grid parity.

While this is sort of a PR/promotional piece, the numbers in the report are backed up by my real world experiences. As long as the world doesn’t fall apart anytime soon, solar power is on track.

[Edit 6/16: Updated link to Sunpower LCOE paper after their website redesign]

The first, and biggest thing, is a reduction across the board, reducing energy consumption by 33%. From air conditioning and heating, lighting, cooking, refrigeration, and entertainment.

If we figure a 1% per year efficiency increase every year in terms of air conditioners, heaters, refrigeration, etc, we’ll see a 22% decrease in power consumption compared to 2008 usage.

For things like lighting, entertainment, computing, we tend to see big steps rather than small ones. LCDs replaced CRTs – old 20″ CRTs used 100-150W, my 24″ LCD uses 65W. The next incremental upgrade is the LED-based LCDs, however power consumption can vary – some LED-based LCDs have power consumption higher than their traditional LCD counterpart, though most have lower power consumption, which is the drive to use them in laptops to save more power. The next major increase in efficiency we will hopefully see Organic LED (OLED) based displays start to show up in the next few years (though we’ve heard that for the last three or four years now). As OLED screen sizes start to increase and prices come down, larger flat panel displays will be available, driving down the energy consumption for home entertainment.

From there, OLEDs could even infiltrate lighting if they can manage to produce white OLEDs that have a long enough lifetime to further bring down the usage of lighting as a percentage of total electricity consumption.

Likewise, personal computing is getting smaller. While my 486 chip ran cool enough not to need a heatsink, my current quad core Q9450@3.2GHz requires a big heatsink and fan, and sucks a lot of power when at full load. But not everyone needs that speedy dual core. Intel has done a lot to reduce power consumption when idle, starting in notebooks back in the early Centrino days, and continuing today across all lines – servers, desktops and laptops. The days of a 2.4GHz dual core Intel Atom aren’t far away (2011?), and as for email, web, movies and video, that’s all people might need. Enthusiasts and gamers will always buy the fastest and the best, but that part of the market shrinks as more gamers move to consoles as well as the general movement away from power hugrier desktops to more efficient laptops.

Next up is the new energy demand on the grid – specifically plug-in electric hybrid vehicles, like the Chevy Volt. Google estimates they’ll be 42% of the US passenger vehicle fleet (also known as LDV or light duty vehicles) by 2030. Is it possible? Yes, but it requires PHEVs supplant 90% of LDV sales, which is feasible, if not higher – the only reason I can think of not buying an PHEV would be for the very cheap cars (sub $10K) and to those who live in very extreme rural environments where electricity is scarce (those who live off the grid) and its not worth it to use electricity instead of gas.

When it comes to energy consumption, I think Google is on track, and their plan is feasible. We will see increases in efficiency in things like household appliances, improvements in household lighting, and technology will provide for decreases in entertainment and technology. However, stand-by power is an important factor, and should be addressed as well (it doesn’t appear in their knol page about their plan).

Now we’re on to energy sources. The two biggest up and coming renewable sources are wind and solar, however, geothermal is lurking in the dark. Both have their obvious limits on usage. The issue from there is what to do about load matching, and how much backup capacity will be needed.

Lets start with solar. I’ve said before that I estimate grid parity will arrive sometime around the time that the freshly passed solar energy ITC expires at the end of 2017. Google is predicting 250GW worth of solar power installations, and an approximate 500TWh/yr worth of production (a capacity factor of 22%, which is entirely appropriate). Their forecasts show 66GW of domestic capacity being installed between 2010 and 2020, which might sound unreasonable – an average 6.6GW installed per year – but current research estimates that worldwide, there will be between 50 and 80GW of panels manufactured per year by 2015. I would venture to say that it would be more difficult to find the installers and construction workers required to install all those panels, whether they be in large or utility scale projects, or rooftop installations for homes and businesses. And nevermind the credit crunch when it comes to financing the projects. Recently, two leading solar power companies had their stocks downgraded because of expected oversupply in light of the worldwide economy and credit situation.

Next is wind, which is estimated at 380GW by 2030. The US DOE had done studies for 300GW by 2030, and they found it possible with a few minor issues. Google assigns a 30% capacity factor to the wind turbines, to generate approximately 1000TWh/yr. This seems slightly high, but assuming that turbine installers stick to the highest wind speed areas in the US and then sends the electricity to urban areas via 500kV lines, it seems feasible. A group of utilities have formed the Utility Wind Integration Group, who’s goal is to advance the technology and application of wind into the power grid. In a study from November 2003 they stated that,“…the results to date also lay to rest one of the major concerns often expressed about wind power: that a wind plant would need to be backed up with and equal amount of dispatchable generation. It is now clear that, even at moderate wind penetrations, the need for additional generation to compensate for wind variations is substantially less than one-for-one and is often closer to zero.” This bodes well for wind, since it will be possible to integrate wind into the grid without having to build a large standby capacity in case the wind decides to not blow.

Finally, with geothermal, Google invested $10M USD into Enhanced Geothermal systems. The difference between traditional geothermal systems, of which the US estimates there is a total of about 10GW worth of recoverable energy from traditional geothermal, is that with the EGS, you fracture hot rocks and pump water down into the earths surface, instead of depending on naturally occurring sources of water to be in the right place to become heated enough to turn into steam. Google predicts that of the 80GW of geothermal production in the US, 65GW will be from Enhanced Geothermal systems, and the total TWh produced from all Geothermal systems would exceed that of solar power because of the much higher capacity factor of Geothermal systems (between 89-97% CF, compared to solar’s 22% and wind’s 30% CF).

Still, the ultimate issue with solar and wind is availability – how do we use the power its putting out without a massive amount of overbuilding and tons of power lines to keep everything balanced between production regions and consuming regions. Discussions are up for everything from large scale batteries and Vehicle-to-Grid technology to potenial energy storage systems (water reservioirs, flywheels, etc). Google estimates this could add $20/MWh, compared to the pricing range between $35/MWh for off-peak power and up to $200/MWh for peak power under the highest demand periods (during the summer when the need for power for air conditioners is highest). While nuclear and geothermal power could be used for part of the base, wind  would also need to make up a part the total baseload. Solar (PV and concentrated with auxiliary thermal storage) would be well suited for matching peak demand, as well as shaving down the base usage during the 9 months out of the year its not summer.

The key to Google Energy plan energy efficiency as well as smoothing out the generation of intermittent sources like wind and solar. I’m confident the prices of wind and solar will come down enough and efficiency will go up, as we figure out how to make the products more robust, reduce overhead as production scales up, and material substitution to reduce the cost of parts.