Judging from industry hype, it might seem that the smart grid will solve virtually all of our problems relating to energy, transportation, and the economy moving forward. Smart meters, distribution management automation, and other smart grid technologies will not only reduce both customer and utility costs and optimize the power grid akin to an Internet of Energy, but also is portrayed as vital to efforts to increase renewable energy production.
Last month, I attended the “Wind and Solar Integration Summit” in Scottsdale, Arizona, as a starting point for my research on a forthcoming Pike Research report. There was plenty of discussion about wind and solar forecasting, different types of energy storage, and the different challenges that face regional grid operations across the United States. Interestingly, I rarely heard the term “smart grid.”
Part of that, no doubt, is because the focus of efforts to date on integrating variable wind and solar power has been at the wholesale, transmission level of grid service, instead of at the distribution level, where smart grids, microgrids and virtual power plants are absolutely vital for integration. It’s at the wholesale level where the money is right now, integrating bulk renewable energy into so-called organized markets managed by entities known as independent system operators (ISOs).
The summit did provide some good data points, among them the fact that wind integration costs
generally run from $3 to $12 per megawatt hour (MWh), which at today’s wind penetration levels adds up to $270 million to $1 billion in just the United States. Less data is available about solar integration costs since utility scale solar PV is a rather recent phenomenon, but one can assume roughly the same order of magnitude.
Iberdrola, the Spanish operator, has more than 3 gigawatts (GW) of wind power capacity in current operation in the Pacific Northwest. The company is among the leaders in investigating how better forecasting can reduce integration costs. According to the company, so-called “day ahead” forecasts are already about as accurate as they can get, with error rates ranging from zero to as high as 18% for Iberdrola in the Bonneville Power Administration’s (BPA) grid control area spanning Washington, Oregon, Idaho and Montana. (The equivalent forecasting error rate for day ahead forecasts in Europe is closer to just 5%, reflecting, perhaps, a more mature technology/policy integration.)
The real challenge for wind and solar forecasting is in the “hour ahead” and “intra-hour” data. Within this forecasting envelope, error rates can exceed 30% for wind power. The shorter the scheduling interval – e.g., every five minutes, as is the case in Texas – the more accurate the forecast. (This is one reason why BPA has struggled in the past is that it used to only schedule wind hourly, and even today schedules wind power every 30 to 60 minutes).
Which variable renewable energy technology offers the greatest integration challenge? While wind power is less predictable than solar power, the output from the utility scale solar PV project can ramp down instantaneously with cloud cover. In contrast, wind turbine ramps tend to be more gradual due to spinning machinery.
Beyond forecasting, the most heated discussions at the Summit pertained to energy storage. It became clear that the perception that energy storage was too expensive may not always be true. Energy storage is not a monolithic resource, but rather an emerging grouping of technologies that can offer long-term and short-term solutions for variable renewable resources. The cost of a flywheel providing frequency regulation is a completely different animal than a compressed air storage unit offering long-term energy storage. The storage firm A123, working with AES Storage, has bragging rights to a 32MW storage project offering frequency regulation services in the Pennsylvania-New Jersey-Maryland (PJM) grid control area today, as well as a 12MW spinning reserve service project in Chile, South America.
The most provocative take away from the Scottsdale conference was a recently released study by Alstom Grid that surveys the world about solutions to the challenges of wind integration. This report actually does reference the smart grid, highlighting the role of demand response, dynamic line ratings and transformer load management as keys to moving forward with planned wind project integration throughout the globe.
The truth of the matter is that the integration of renewables is not a reliability issue, as these resources are being integrated around the world without a smart grid. It’s really all a matter of costs to ratepayers. The far larger challenge is at the distribution level, which is where microgrids and virtual power plants come in. I’ll have more on that topic in a future blog post.