In early July, just as a light rain began to fall, I headed to the bank on my bike. Given the rarity of summer precipitation in the Central Valley, I didn’t bother to take a rain jacket. The summer sprinkle turned into a heavy downpour. While biking as fast as I could back to the office, I got soaked to the bone. Miles away, the rain also drenched more than 300 MW of solar power. The sun supply stopped in its tracks during the July 3 monsoon. Unlike me on my bike, the grid operator was prepared. The demand for more renewable power in California and its successful integration has led the grid operator to invest heavily in the latest weather forecasting technology in its impressive new headquarters in Folsom. When spot on, more sophisticated weather forecasting tools enable the California Independent System Operator to buy alternative power supplies in the more affordable day-ahead market to replace sudden drops in power caused by clouds, high winds, precipitation, or other predicted weather events. CAISO had electricity supplies lined up to fill the void left by the 65 percent drop in solar power caused by the change in weather July 3. Earlier, the grid operator had to make up an 800 MW loss of wind power when powerful Santa Ana winds hit Southern California. CAISO lined up resources to meet that renewable energy shortfall. Improved weather forecasting technology helps the grid operator avoid getting caught off guard and better predict how Mother Nature will impact wind and solar power supplies, Jim McIntosh, CAISO director and executive operations advisor, told me. Better forecasting is not only good for reliability, but is expected to help lower the cost of alternative power supplies. Improved weather forecasting lessens the need for fossil-fueled generation back up for intermittent renewable power facilities. If the weather guys and gals are correct, fossil plants will run less and make way for less expensive, carbon-lite intermittent resources--except when it rains and Santa Anas hit. Currently, in Spain, which has high levels of wind and solar power, each megawatt of renewable energy is matched with equal amounts of fossil-fired back up generation, driving up the costs of growing cleaner power supplies. Perhaps they’re still using necromancers instead of the National Weather Service, and are less able to predict when it’s going to rain in Spain. There is no set fossil back up requirement in California, although the amount of operating reserves is doubled for renewable supplies, McIntosh said. It was impressive to listen to McIntosh talk about improved weather forecasting tools, particularly while taking in the huge colorful, complicated displays in CAISO’s control room. Mounted on the wall in the grid operator’s nerve center are massive bright screens monitoring and displaying a variety of changing real time and projected conditions, from cloud cover to amount of solar and wind output, location, and current and expected load. The information comes from several sources, including telemetry at project sites, Google Earth Data, weather forecasting inputs, as well as high-end geospatial technology that integrates and layers several factors, including barometric pressure, GPS readings, temperatures, and fire threats. In spite of all the concerns about integrating higher levels of renewables inside and outside CAISO McIntosh worries most about fires. They can severely damage major transmission lines, making a drop of 300 MW of solar look like drizzle. I couldn’t, however, quite forget that weather forecasts are notoriously unreliable. If the grid operator is off, it can cause more than a drenching. CAISO has steadily erred on the side of costly caution--supporting overbuilding fossil facilities--but that could be balanced out with more accurate forecasting. That also would put some downward pressure on rising rates. McIntosh said the grid operator’s forecasting technology accuracy has significantly improved. Its accuracy rate has increased from 30 percent to near 70 percent. “We have to improve our tools,” he added while we looked down at the new, fancy-schmancy CAISO control room. The tools he is on the look out for are expected to lead to better forecasts, improved and increased data, and more accurate models. But it will be several years, and a lot more investment, before the grid operator’s renewables weather forecasting is as accurate as its load forecasts. Tackling the last 30 percent of inaccuracies will be the most challenging. The forecasting accuracy gap, combined with the expected surge in wind and solar power in response to the state’s 33 percent renewable mandate, makes McIntosh and others nervous Nellies. Consider that over the next few years, renewable output could swing by thousands of megawatts. Losing 4,000 MW of intermittent power would be unprecedented, “equal to losing two large power plants within an hour,” McIntosh noted. More accurate and sophisticated data is expected to be gathered by synchrophasor projects in the works. They produce far more data--400,000 pieces of information every four seconds. Not only are they expected to provide a far more accurate look at what is going on with the power system, they allow corrective action to balance the grid within minutes. Given the fast pace of forecasting technology advancements, I asked McIntosh how he planned to deal with constant technological upgrades. He said under the deals CAISO signed, upgrades are included. While I remain unconvinced that weather forecasting will ever be a reliable science, I may give the CAISO a ring next time I head out on my bike during a light rain.