No “magic number” for renewable power

Michael Giberson

Peter Behr, reporting for ClimateWire in an article online at the New York Times, captures some of the discussion surrounding the recent NERC report on integrating renewable power to the transmission grid [NERC Press Release] [NERC Executive Summary] [Full NERC Report].

The vast expansion of wind and solar power planned by the Obama administration and congressional leaders is fraught with challenges for the nation’s aged electricity network, grid monitors with the North American Electric Reliability Corp. say.

But a NERC report released today does not call for a slowdown in deployment of renewable energy. Officials expressed confidence that technology solutions will arrive in time.

A section heading in the article states: “No one knows the ‘magic number’ of renewable capacity.”

Revis James, who directs energy technology assessment for the Electric Power Research Institute, said that a critical question hangs over the push to increase renewable energy output. “How much renewable energy can you have before [you] have to have systemic improvements to the system to handle the variability of renewables?” he said. “Is 10 percent too high? No one knows what the magic number is.”

Let’s get right to the answer: there is no “magic number.”

Understanding of the relationships between renewable energy and grid operations is growing with experience, but that understanding is not ever going to yield a magic number. Instead, there are multiple relationships in play and many margins of analysis.

Consider, for example, the nature of the grid that the variable resource is attaching to. ERCOT got lucky in that restructuring of the industry in Texas led to a lot of investment in efficient, flexible natural gas generating plants. As increasing amounts of wind came onto the ERCOT system, there was already a lot of new, efficient, complementary gas-fired generation around to help manage the variable output. Had economics favored large coal-fired steam or new nuclear units in the early 2000s, ERCOT would have had much more difficulty accommodating wind.

For another kind of example, consider the effects of changing relative fuel costs on the dispatch order for a utility or regional power market.  As the recent FERC “State of the Market” report explains, as gas prices fell faster than coal prices in late 2008, efficient gas units became competitive with and even cheaper than baseload coal in some areas. But if flexible efficient natural gas units are operating as baseload units – so near the top of their output range most of the time – then they have little flexibility available to follow load swings or compensate for variable wind power supplies.

Ancillary services practices matter, too. Regional grids with relatively open, transparent balancing markets used for redispatch will find it easier to accommodate variable wind power than systems that rely upon heavy penalties for “unscheduled energy” and cumbersome administrative congestion management procedures. Similarly, grids that have economical and flexible means for procuring regulation service (also called “automated generation control”) and responsive reserves with do better than grids with rigid tools for obtaining such ancillary services. Both of these elements have tended to favor RTO/ISO type markets for the integration of renewable power over traditional, vertically-integrated utilities.

One of the challenges of power market design for RTOs has been figuring out how to pay and how much to pay for generators (and load) willing and able to provide flexibility to the system operator. Frequently, system operators simply assumed that whatever flexibility a generator had should be and would be made available to the system operator to use for reliability purposes. While interruptible loads were typically paid for the service the provided, the programs and the prices were not initially well-integrated to market operations. Over time, the system operators, market participants, and regulators are learning that you get what you pay for. When the RTO didn’t pay for flexibility, it tended to see less and less of it made available.

If the system operator needs flexibility to manage the system reliably, the market operator needs to be able to pay for it (and, to complete the idea, needs to be able to charge the generation owner or load whose system use requires the presence of other, flexible units on the system).

At the recent Gulf Coast Power Association meetings, some representatives of generation and financial investor interests recommended ERCOT pursue a capacity market to support investment in generation. (I didn’t hear any load-side representatives endorse the capacity market concept.) The panel on renewable energy was asked whether a capacity market might be needed to support investment into sufficiently flexible stand-by generation.

The answer here is no, a capacity market is not needed to support investment in complementary generation units or responsive load, but the regional spot market better have some other means for paying for the necessary flexibility. Well-designed ancillary services markets and complementary dynamic procurement practices should do the trick.

Only it is no trick, just a matter of working out the market design.