Renewable electric prices fall while challenges riseBeyond policy goals, the growth of renewables is supported by their improving cost outlook. According to Lazard’s year-end 2017 estimate, levelized cost of energy (LCOE) for utility-scale renewable electricity continues to fall, averaging $45 per megawatt-hour (MWh) for unsubsidized wind power and $45 to $50 per MWh for utility-scale solar, compared to approximately $60 per MWh for combined-cycle natural gas.8In several geographies, solar and wind are already competitive with other sources of generation based on LCOE, even without tax or production subsidies. In LCOE terms, these resources are expected to be the cheapest source of electricity within the next decade
We believe significant steps can be taken toward decarbonizing the electricity supply through thoughtful, concerted action, as we discuss below.
High renewable penetration can cause several issues in the operation of the grid that can vary by geography, depending on, among other things, the mix of renewable-energy sources (solar versus wind), the availability of transmission and distribution (T&D) capacity, the fleet of nonrenewable generating stations, and the shape of electricity demand. There is no universal, one-size-fits-all solution to integrating ever-greater amounts of renewable generation into the grid. What works in Philadelphia may not work in Portland or Phoenix.
Heading off current and future challenges is not simply a matter of tinkering at the margins with market rules or mandating a set level of electricity-storage projects. Holistic solutions that encompass supply, demand, regulation, and market structure are needed. Storage could be part of the solution, as could supply-side resources, customer programs, and regulatory leadership. The solutions could also recognize that different resources provide different services and thus a differentiated set of market products and customer programs will likely deliver a lower cost solution.
Operating electric grids with high penetration of intermittent resources poses unique challenges for utilities and grid operators. While not limited to California, that state’s much discussed “duck curve” (Exhibit 1) illustrates one of the difficulties managing a grid with a high percentage of renewable generation, which sometimes is dispatched at zero or even negative variable cost, resulting in displacement of other resources which may be needed to manage flexibility.
Though this particular iteration of the “duck curve” is a projection, it differs little from recent years’ operational facts. During the hours of a spring day when the sun is shining most brightly, that is, from 9:00 a.m. to 3:00 p.m., renewable generation surges, displacing other forms of electric generation. The problem is most acute in the first and second quarters of a year. During the first quarter of 2017, the price of generation on the CAISO fell to zero or less than zero for as much as 15 percent of the time between the hours of 11:00 a.m. and 4:00 p.m.9
As solar generation peaks midday, non-solar generation is ramped down. More importantly, later in the day those ramped-down plants will need to ramp up as solar output declines. The significant amount of renewable output during the midday solar peak means that nonrenewable-generation plants will have to operate close or at their minimum generation levels and possibly shut down for several hours each day, stressing equipment as well as operating economics. For wind-based systems, the need for flexibility might manifest in different ways. For example, significant investments might be needed to alleviate power-export constraints from zones with significant wind potential to major load centers, an effect that has already been observed in Texas’s Panhandle Renewable Energy Zone. Additionally, nighttime wind-overgeneration issues might arise, an effect already apparent in the Electric Reliability Council of Texas and Southwest Power Pool. Furthermore, short-term wind-forecasting errors might hamper the grid operator’s ability to match supply and demand, degrading the system’s primary and secondary frequency response, as has already been observed in several European markets