Finding the Balance

Electricity is not like other commodities, partly because, at least in our part of the world, it is very expensive to store it. There are no warehouses or tanks full of electricity ready and waiting for someone to switch on a light. As a result, the entire grid has to respond when you switch on your light, all the way back to the generating plants. Automated equipment handles small fluctuations, but generators must be started up or shut down to handle larger changes in demand.

The situation is different with natural gas. The U.S. stores large quantities of natural gas in tanks and caverns — roughly a 2-month supply in the summer, down to about 20 days in the winter. This past winter the inventory got down to around 10 days a couple of times.

If you have natural gas service, you have both a meter and a pressure regulator between your house and the gas line. This allows the pressure in the line to vary over a considerable range, while keeping the pressure for your stove and furnace fairly constant. There’s also a meter on the side of your house to measure your electricity use, but there’s no regulator. It is up to the grid operators to keep the voltage and frequency within very tight tolerances at all times. Relatively small deviations are not only a problem for your clocks and light bulbs; they can cause considerable damage to transformers and other parts of the electricity infrastructure.

In our part of the country, electrical demand can vary over a 3-to-1 range across the year and sometimes over just a day or two. The highest demand is typically on hot summer days, while the lowest is late at night in mild weather. The efficiency and cost of running a large power plant is lowest when it is either producing at full capacity, or not running at all. Summer demand peaks typically last a few hours at most, while large coal plants take several hours to start up and shut down. Nuclear plants take much longer – often a few days. In order to handle the large variation in demand, there are several different types of generators connected to the grid. Large nuclear and coal plants are generally the lowest cost (along with wind, more on that later), but cannot be started up and shut down quickly. At the other end of the spectrum are gas-turbine “peaking plants” that can be started up and shut down in a few minutes. Because they are not used very often and are relatively inefficient, they produce energy at a much higher cost than the large “baseload” plants.

Renewable energy

What about wind turbines and solar panels? Fuel cost is zero, and carbon emissions and other pollution related to constructing them is minimal. The cost of wind power has come down over the last ten years, and is now competitive with coal and natural gas; the cost of solar is competitive in desert areas, and it is dropping fast. However, wind turbines produce energy when the wind blows and solar panels produce when the sun shines. They cannot be turned on and off to meet demand. They are somewhat predictable in the short term, but over longer periods they are about as predictable as the weather. Variability is reduced by spreading them across the countryside, reducing the impact of local weather variations.

Of course, there are renewable energy sources that can run continuously, just as coal and nuclear plants do. These include geothermal and hydropower, along with biogas (methane digesters on dairy farms and sewage plants, for example) and biomass (crop waste, wood chips, etc.). Norway and Iceland are largely powered by hydro and geothermal plants, respectively.

If renewables are to become a major component of our energy mix, we must find ways to manage them that preserves the stability and reliability of the grid we all depend on. One way is to match them up with natural gas plants that can be turned up or down much more quickly than coal or nuclear plants. If there is enough wind and/or solar to at least sometimes supply the total demand, then it can become almost impossible to use large coal or nuclear plants, because they cannot react quickly enough to “fill in the gaps”. This situation is fast approaching in California, and is already happening at times in Denmark. The solution in Denmark has been to partner with Norway and Sweden, both of whom get most of their energy from hydropower. When the wind is blowing in Denmark, they ship the surplus up to Norway and Sweden, who then slow down their hydro production and let the reservoirs build up, or in some cases even pump water back uphill. When the wind slows down, Denmark imports hydro power from Norway and Sweden.

Another solution, of course, is to add batteries of one sort or another. California has mandated a significant buildup of electrical storage capacity over the next six years. Currently, batteries cost around $400/kWh, making them more expensive than gas-powered reserve capacity, but the cost is expected to drop quickly over the next few years as electric vehicles become more popular and new battery technologies come online.

Demand Response

Why, you might ask, can we not adjust demand to match the available supply, instead of adjusting supply to meet demand? Part of the answer is that there is lots of opportunity to adjust demand, but we lack the infrastructure for managing it in a way that is acceptable to energy users. Our existing electric grid was designed and built over the last 100 years with the assumption that a few large power plants (the supply side) would supply power for large numbers of mostly small customers (the demand side), and that there was no practical way to adjust demand to meet supply. When  we measure usage over months rather than hours, it is not possible to ask people to pay more when demand is high, and existing “off-peak” rates primarily avoid buying high-cost energy during demand peaks, rather than adjusting demand directly to offset the need for expensive peaking power. Our current measurement and control mechanisms are simply not adequate to manage demand in the same way we can manage power plants.

There is considerable work going on to adapt energy demand to supply. One promising approach is to treat thermal energy storage devices, like water heaters, as though they were batteries. Cold-storage warehouses can also be used as batteries.

Of course, if people start buying electric cars, there will be lots of batteries plugged into the grid that can not only run the cars but also respond to the availability of energy in the grid.

Energy productivity

Of course, we have to balance our energy supply with our energy needs over periods of years, not just minutes and hours. Traditional power plants can take 10 years or more to plan, permit, build, and bring online. They represent major capital commitments, and they are expected to produce and sell power for 40 years or more. Companies (or cooperatives) who make such investments can be in trouble if the demand does not materialize. Some companies in Europe are in big trouble because they have invested heavily in new capacity that will probably never be profitable. Partly this is because during the 10-year plan-build cycle, a lot of solar and wind came online and changed the dynamics of the market dramatically. But partly it’s because the European economy has become more energy-efficient in the meantime.

Conventional wisdom says that a higher standard of living is directly tied to using more energy, but actually the relationship between energy consumption and economic growth or standard of living is indirect at best. What we want is better lives, and when we see the soaring price of filling our tanks, or involvement in the Middle East because that’s where the oil is, it sure looks like we have an energy problem. But according to McKinsey, there is plenty of room for economic growth without consuming more energy; in many cases, we would be better off investing in energy productivity than in new energy sources. Indeed, the U.S. Department of Energy says we can double our energy productivity by 2030, and do it in ways that actually contribute to economic growth. In other words, Negawatts are cheaper than Megawatts.

Achieving sustainability

In the long run, we have to find a way to support our civilization on energy that does not involve digging up increasingly limited supplies of fossil carbon or destroying our planet. It seems clear that a combination of renewable energy, storage, demand response, and improved energy productivity can meet all our needs, although in the near future this will be easier in the U.S. than in Europe. In the short term, we may need additional gas-fired capacity that can fill in the gaps when solar and wind are in short supply. In the much longer term, we may learn how to harness fusion energy, but that promise has stubbornly stayed 50 years away for at least the last 50 years. Does this mean it’s time to stop building coal plants? I believe the answer is “Yes.” It’s very hard to believe that a new coal plant today could ever pay back its investment, and it’s clear that once we start accounting for the external costs of mining and burning coal, the cost will be higher than the combination of solar, wind, geothermal, biogas, biomass, storage, and demand response. Meanwhile, the cost of wind, solar, and storage continues to fall. This is the future.