Moving forward with the next few years might seem straightforward enough, implement efficiency programs and purchase a little renewable energy.

But what happens in 20 years or even 50 years down the line, when utilities will be expected to reduce their emissions some 50 percent or more? To stabilize the climate, experts are calling for an 80 percent reduction in carbon emissions, below 1990 levels, by 2050.

RMI has spent time modeling these future low-CO2 scenarios from a utility's perspective. Specifically, we've created a simplified electric dispatch model that utilizes real demand data, renewable supply data, and technology characteristics of demand response and electrified vehicles to determine how various supply and demand resources will interact at the hourly level. This allows us to quantify the resource adequacy, cost, reliability, and CO2 footprint of what we call the "next generation" utility.

So far, we've found that a low-cost and highly reliable electricity resource plan will necessarily include a mix of energy efficiency, distributed generation, renewable energy, demand response, and electric storage -- largely from electric vehicles.

Here are some specifics on our findings, including our recommendations for how utilities can mitigate the variability of renewable technologies.


The "Easy" Part: Efficiency and Distributed Generation

Efficiency: Reducing the supply load nearly always presents the cheapest way to "provide" energy. For example, improving the efficiency of commercial lighting lowers the lighting load, and it also lowers the commercial cooling load because the air conditioning unit doesn't need to work as hard to remove waste heat from lighting. Cost savings are well-documented. In California, the price of efficiency programs has averaged 2 to 3 cents per avoided kilowatt-hour -- about one-fifth the cost of electricity generated from new fossil-fuel plants.

Distributed generation (DG) -- like rooftop solar and microturbines -- can cost-effectively reduce CO2 emissions, especially when used in combined heat and power applications. Because DG is located at the load, it doesn't suffer transmission & distribution losses, and its waste heat (if it has any) can be captured to offset heating loads. Distributed generation is used for building- or industry- specific purposes, so it is not dispatched like central generators. Instead, it can follow the electric or heating load of the associated facility.

Firm Renewables

Firm renewable energy resources include hydroelectric, geothermal, and biomass power. The potential generation for these sources varies with geography. In our model, firm renewable generators act much like traditional generators. Unlike traditional sources, however, utilities will need to be careful not to rely on more capacity than is available in the location being modeled, and to consider the status of the technology for a given year.

How to Incorporate Variable Renewables

The challenging part for utilities could be managing the variability of renewable resources like solar and wind. Fortunately, and contrary to conventional thought, numerous strategies exist. Utilities have always dealt with variations in demand; consumers use more electricity during the day than night, for example. Now, they will have to plan for variations in supply and demand.

Utilities can mitigate the variability of renewable generators through a variety of strategies, including geographic distribution, shifting demand profiles, and adding energy storage.

Geographic Distribution. Because the wind doesn't blow at the same time everywhere, combining multiple wind sites yields a less variable output than one site alone. The same is true of solar sites. Generally, the farther apart the sites, the less variable the combined output.

In most electric systems, high penetrations of renewable energy means large amounts of variable wind and solar resources. In conventional wisdom, these variable resources require backup from quick-ramping fossil fuel generators. Demand response, however, provides a better alternative because quick-ramping generators are inefficient and expensive to run.

Demand-Response. Rather than dispatching elastic supply resources to meet an inelastic demand, demand response makes the demand elastic - shifting it to times when renewable generators are producing power. Demand response can take two forms: peak-reduction and dynamic load shifting. Peak reduction is the conventional form of demand response and can be useful in displacing peaking generators. Dynamic load shifting is a more advanced form of demand response that relies on SmartGrid technology to shift flexible loads (e.g. washing machines, dishwashers, electric vehicle charging) to times when renewable energy is abundant.

Electricity Storage. Electricity storage, including pumped hydroelectric, compressed air, and batteries, provides another strategy for mitigating the variability of wind and solar resources. Historically, electricity storage has proved prohibitively expensive. However, the combination of electric drive vehicles and the SmartGrid (now receiving a great deal of government investment) provides a way for utilities to take advantage of battery storage without bearing the full cost of the batteries.

Whenever an electric vehicle is plugged in, its battery could be available to soak up excess renewable power or to provide stored power when needed. Of course, the battery must also be used for driving, so limits are placed on charging and discharging. In aggregate, electric drive vehicles increase the electric load, but they also provide inexpensive energy storage and can therefore enable larger penetrations of renewables.

So this is how we see utilities working 30, 40, and 50 years down the line. We are partnering with electric utilities and regulatory bodies to refine the results of our research and apply it to utility-specific situations (for more information, see our Next Generation Utility page). But it's exciting to finally see, on a realistic hourly level, what the next generation utility could look like.

Josh Traube is an analyst with Rocky Mountain Institute's Energy and Resources Team. He specializes in process efficiency for industrial clients and computer modeling for electric utilities and RMI research.

Photo CC-licensed by Flickr user pfala.