Solar and wind are coming. And the power sector isn’t ready.

The rise in renewable energy will scramble the decision making of grid managers.

The US electricity system is at an extremely sensitive and uncertain juncture. More and more indicators point toward a future in which wind and solar power play a large role. But that future is not locked in. It still depends in large part on policies and economics that, while moving in the right direction, aren’t there yet.

And so the people who manage US electricity markets and infrastructure, who must make decisions with 20-, 30-, even 50-year consequences, are stuck making high-stakes bets in a haze of uncertainty.

That uncertainty has increased markedly under the recent Republican administration (somewhat ironically, given its oft-stated goal of “regulatory certainty”). Under President Obama, the feds established a consistent cross-agency push toward clean energy. The long-term trajectory was clear.

Now it’s been thrown into doubt. President Trump has embraced fossil fuels, and the owners of struggling coal plants are appealing to the Federal Energy Regulatory Commission (FERC) for bailouts.

Should utilities and market managers bet that the Trumpian revolt against modernity will succeed in slowing the growth of renewable energy? Or should they bet that it’s a passing phase and renewable energy will triumph?

A fascinating bit of new research from the energy geeks at Lawrence Berkeley National Lab (LBNL) sheds some light on the stakes involved.

Shutterstock

In a nutshell, things will look different in an electricity system with lots of variable renewable energy (VRE) — different prices, a different shape of demand, different timing, different needs — and if the people managing the electricity system bet on low VRE and get high, they are going to screw up all sorts of things.

If the US gets serious about renewables, the electricity system will look very different

As of 2016, wind and solar power — VRE — provide 7.1 percent of US electricity. VRE affects utility and market decisions, but it is not yet central to them. The LBNL team (Joachim Seel, Andrew Mills, and Ryan Wiser) notes that “many long-lasting decisions for supply- and demand-side electricity infrastructure and programs are based on historical observations or assume a business-as-usual future with low shares of VRE.”

But what if VRE takes off? What if it hits 40 or 50 percent of national electricity supply by 2030? (Climate hawks would prefer an entirely decarbonized power sector by then; neither goal will be possible without a serious national policy push.) Would high VRE penetrations substantially change the decisions that energy regulators, policymakers, and investors need to make?

In a word, yes. They would.

The team modeled the effects of high (40 percent) VRE and found several notable changes relevant to the operation of wholesale energy markets. Here they are all at once, in a giant, info-packed chart!

LBNL

Now that your eyes are bleeding, let’s back up and walk through the changes.

The team modeled four 2030 scenarios: a baseline, with VRE shares frozen at 2016 levels, and three high-VRE scenarios, one that’s wind at 30 percent share and solar at 10 percent, one that’s the reverse, and a “balanced” 20-20 scenario. They ran these four scenarios for each of four energy markets in the US: the Southwest Power Pool (SPP) covering Kansas, Oklahoma, and portions of surrounding states; the New York Independent System Operator (NYISO), the California Independent System Operator (CAISO), and the Electric Reliability Council of Texas (ERCOT).

VRE will change prices and dynamics in energy markets

Here are some of the results, which will throw wholesale markets into a new equilibrium.

1) VRE reduces average wholesale power prices.

In all high-VRE scenarios, in all markets, average wholesale power prices go down. Depending on the scenario and region, the drop is anywhere from $5 to $16.

LBNL

Note that average prices fall the most under the high-solar scenario, in every market but ERCOT. Unlike the other states, Texas is a bit isolated, running its own grid with few interconnections to other grids through which it can import or export power. It has to deal with all that solar on its own (more on that later).

Lower prices are good for consumers but bad for owners of big, uneconomic coal and nuclear plants, who rely on high prices to keep running. (Yes, it is a peculiar market in which most of the people responsible, including the president, view low prices as a threat.)

2) VRE bumps fossil fuels off the grid.

In high-VRE scenarios, markets see anywhere from 4 to 16 percent retirement in “firm capacity,” i.e., coal, oil, and steam turbines. The exception is CAISO, which sees a small, 2 to 4 percent boost in firm capacity via the growth of natural gas. (Natural gas also grows in SPP and NYISO, though it’s offset by coal and oil retirements.)

Notably, VRE reduces the amount of energy generated from fossil fuels (MWh) much faster than it reduces capacity (MW), anywhere from 25 to 50 percent (the most in NYISO). Basically, every new kWh from VRE displaces a kWh from fossil fuels.

3) VRE makes periods of very low prices and very low emissions more frequent.

Depending on the market and scenario, high VRE buildout reduces overall carbon emissions anywhere from 21 to 47 percent and “leads to an increase in frequency of hours with very low marginal emission rates ranging from 5% of all hours in CAISO (wind scenario) to 31% in SPP (solar scenario).”

Also more frequent under high VRE are periods in which wholesale power prices are extremely low, under $5 a MWh. (It’s these periods that so wreck the economics of big coal and nuclear plants.) The effect is especially pronounced in ERCOT under high solar.

LBNL

4) VRE changes the shape of daily demand …

This is the most interesting bit for the energy nerds. I and many others have written about the “duck curve” that shows up in the shape of diurnal (24-hour) demand in California as the share of solar power increases — you can see it on the top left in the chart below. LBNL’s research has revealed what shapes would appear in the demand curves of other regions in high-VRE scenarios.

LBNL

As you can see, the duck effect is most pronounced in ERCOT, which has the least ability to export excess solar power (and ends up doing the most curtailment).

5) … and pushes demand peaks later in the day.

VRE doesn’t have much effect on the timing of peak demand in CAISO, but in the three other areas, high VRE pushes diurnal peak demand back a few hours (and raises it higher).

LBNL

6) VRE makes prices more volatile.

Power prices in high-VRE scenarios are lower on average, but they move around more. Solar scenarios are more volatile overall, though prices in high-wind scenarios swing over a wider range. In spring in California under a high-wind scenario, “energy prices in the morning may be at zero on some days while prices may reach up to $55/MWh on other days.”

7) VRE makes the services that support it much more valuable.

Along with the volatility of VRE comes a need for more services that provide flexibility and stability to compensate. “Ancillary services” for the grid include things like spinning reserves, frequency and voltage regulation, demand response, and, most notably, storage. Those services will command higher prices under high VRE, especially high solar, drawing more competitors into those markets (a good thing for storage).

So why do all these changes matter?

High VRE scenarios call for different decisions from utilities and regulators

The LBNL team set out to answer a question: How would the changes high VRE brings affect big decisions around electricity policy and infrastructure?

VRE turns out to be material to those decisions in all sorts of ways. In fact, LBNL offers a helpful table with (squints) 11 different kinds of supply- and demand-side electricity decisions, along with how high-VRE scenarios might affect them and how decisions might change in the face of high VRE.

LBNL

I won’t burden you with a comprehensive review. Just a few examples LBNL highlights.

• When considering a portfolio of energy efficiency measures, VRE will make it more important to consider their timing as well as their size.

For example, high shares of solar can depress prices during the day and shift peak times to the early evening. This indicates that traditional on-peak measures, like commercial office building air conditioning programs, may become less valuable while traditional off-peak measures, like street and residential lighting, may increase in value.

• VRE will make it more important to electrify everything, even if, in some contexts (like building heating and cooling), it might mean a short-term reduction in energy efficiency. The more water heaters and electric vehicles are connected to the grid, providing storage and controllable demand, the more stable a high-VRE grid will be.
• If nuclear plants are to survive in a high-VRE scenario (and climate hawks should want them to), they must become must more flexible, capable of ramping up and down in response to swings in VRE. That means “increasing R&D on flexible nuclear plant design and operations, addressing technical regulations on nuclear plant operations, or considering the size of the required incentive (if at all) to either keep nuclear plants operating in a low or high VRE future despite output curtailment, or to increase operational flexibility via plant retrofits.”

This is just a small selection of the kinds of decisions that will need to be made differently if the US is actually going to ramp up renewable energy fast enough to hit its midcentury carbon target.

Right now, the habits and patterns of decision-making shaped by low VRE penetration still have inertia, exacerbated by the lingering doubt Trump has imposed on power markets. But there are many reasons to believe that, Trump or no Trump, VRE numbers are going to keep rising at or faster than their current, already dizzying rates.

The renewable energy future is rapidly becoming the present. Everyone in and around the power sector needs to snap to and get ready for it.