In comments to my previous posts, All Analytics readers have expressed interest in renewable energy sources like solar, wind, and geothermal -- and, naturally, how analytics can help hasten their deployment on today's grids. So let's take a closer look at how analytics can help advance solar energy use.
The amount of energy the sun produces is phenomenal and virtually pollution-free, with zero input costs. The challenge, of course, has been in converting light rays into electricity we can use. The five major obstacles to conversion are cost as compared to fossil fuel plants, site suitability for direct sun, the input costs to make solar panels, the materials needed to create solar panels, and the part-time reliability factor of sunshine hours/seasons.
Analytics can and are playing a role in overcoming this latter obstacle, reliability, by which experts mean that the sun shines only about 12 hours or less each day (depending on where you live and the time of year) and that clouds can at times obscure the sunshine. Because of the sun's nature, solar can only be an additive source of power to fossil or other renewable sources a utility company uses to power its grid.
For reliability, light intensity factor (or irradiance) is the largest stumbling block for utilities right now. In simple terms, irradiance is a function of the wavelengths of light and the ability to convert those wavelengths into meaningful power. Irradiance is difficult to calculate due to the intermittency of solar power -- but calculate utilities it must because they have to blend solar power's intensity, or lack thereof, into their daily planning for generating electricity from other power sources like windmills, coal, or gas plants. And actually, utilities must blend their expectations for solar power intensity into their planning, not just the real-time intensity.
Imagine the enormity of this challenge, with all the related tasks of trying to balance your grid system with mixed solar, gas, coal, and wind sources. Utilities must operate under a sword of Damocles, with the non-compromising expectation that the power they provide is reliable and always available, no matter the time of day or night. Simplified to our solar narrative, that means the utility trying to integrate solar power must always have a gas or coal plant ready to run if a cloudy thunderstorm pops up out of nowhere.
Consider what that means at a utility trying to increase its solar and wind generation.
To help you visualize this, the Solar Electric Power Association (SEPA) offers a fantastic mapping tool showing where current solar investments are occurring. You can get an interactive version online, but note below the large number of utility projects popping up and dominating the map in sunny, cloud-free southern California and Arizona. As great as the weather may usually be in these areas, even they're susceptible to the cloudy day.
To overcome the variable irradiancy problems, utilities deploying solar in these regions are factoring all sorts of calculations into their real-time run management: yesterday's forecast, today's cloud cover, and light intensity, all in real-time, to integrate and simultaneously project the solar arrays' input into the grid operation. They have to, but obviously it's no simple task. In her speech to the Solar Power International 2012 general session, for example, SEPA president and CEO Julia Hamm drew attention to performance monitoring, metering equipment, and forecasting software as technological issues that are still adding substantial uncertainty to the advance of solar power.
Many companies are already trying to rise to the challenge of providing the analytics that can overcome this basic problem clouding solar power's use. But if you talk with anybody in charge of overcoming the challenges at a utility, you'll get an earful. The complexity that's caused by the second-to-second uncertainty of sun power is a daunting hurdle to them.