Electricity Storage and Transmission are naturally complementary, and more of both will be needed. But given limited time and resources, where should those of us who want to see as much renewable electricity on the grid as soon as possible concentrate our efforts? The choice is not immediately clear.
Dennis Ray, ED of Power Systems Engineering Research Center (PSERC) was quoted [pdf, p.11] as saying “Regardless of contractual arrangements that are subject to environmental regulation, the ultimate dispatch pattern that will determine the actual emissions is largely dependent on transmission constraints and reliability considerations.”
Horses for Courses
At a basic level, the preference of transmission over storage will depend on your goals. For those interested in energy self-sufficiency and distributed generation, improved transmission runs contrary to their goals.
The related goal of energy security can cut both ways: while a more integrated national grid might be more vulnerable to large scale blackouts, its greater size would make it less vulnerable to disturbances caused by the loss of any one source of generation. Since transmission can either cause or prevent blackouts, a smarter, more fault tolerant grid seems a better way to combat blackouts than a less connected, more Balkanized grid.
When it comes to goals of increasing the penetration of renewable energy into the grid, electricity storage will likely be essential at high levels of penetration, and new transmission is essential to bring electricity from areas where renewable resources are plentiful to where they are plentiful to population and load centers.
Leaving aside necessary new transmission to bring renewable electricity to market, and electricity storage which will likely be necessary to reach high degrees of grid penetration for renewable electricity, there is considerable scope for both electricity storage and a more robust national grid to make it cheaper and easier to allow quick renewable electricity integration.
Unfortunately, comparing transmission and storage is very much apples and oranges. Storage, in essence, takes electricity produced now and stores if for use later (transfer in time), while transmission takes electricity produced here to where it’s needed there (transfer in space). The relative value of transferring electricity in time and in space depends on the relative price of electricity here and now, to the price of electricity used there and then, as well as the cost in losses from the transfer.
In the balkanized North American grid, we have both daily fluctuation in price and wide geographical price differences, as well as differences in price fluctuations in timing. These variations will make storage relatively more valuable in a location which is far from other parts of the grid, and which sees high daily variation in electricity prices due to variable supply and/or load. A common example of this is that it is often cheaper to build a completely off-grid home if the home is more than a half mile from existing electricity lines.
In contrast, the low cost of connecting to grid service means that no one is envisioning building off-grid homes where electricity service already exists. Note that policymakers are talking about Net Zero Energy Homes (i.e. homes which both import and export electricity from the grid,) not true Zero Energy Homes, which would supply all their own energy all the time.
In a 2005 study "Improving the Value of Wind Generation through Back-up Generation and Storage" from the California Energy Commission (CEC Study) evaluating the use of storage to allow wind to operate as a firm resource found that "even under fairly optimistic assumptions, the energy storage approach is unlikely to perform as well as operating under an Intermittent Resources." If the economics of wind cannot be improved through the direct use of storage, a more robust transmission system will have to achieve benefits at significantly lower costs in order to improve wind economics.
A Simple Example
In order to directly compare the benefits of transmission and storage amid all these variables, I start with a simplified example where the two are more or less comparable. Traveling from West to East is in many ways analogous to traveling forward in time as you cross into new time zones. It’s always an hour later in New York than it is in Chicago.
Let’s assume then that we have a single transmission line connecting two similar areas of the grid, where the marginal costs of electricity have identical patterns throughout the day, but located in time zones an hour apart. Sending electricity from West to East would then have the same economic value as storing the electricity for an hour in the Western grid, and releasing it an hour later. Sending electricity from East to West would have the same economic value as discharging the same amount of electricity from storage, and then re-charging an hour later.
If, for simplicity, we assume that the electricity storage and transmission line have the same capacity, then the transmission line can operate similarly to electricity storage with capacity large enough to charge or discharge discharging continuously for two hours. The two-hour capacity is necessary to compensate for the ability of the transmission line to move power into the past (East to West) as well as the future (West to East)
Still Apples and Oranges?
Some readers may protest that electricity storage has the advantage of charging when electricity costs are low (usually at night), and not discharging until prices peak (usually in the afternoon or evening.) This actually makes less difference than might be expected, because the transmission line can be operating continuously, and the sum of one-hour differences in price will equal the difference between the peak price and the price in the overnight trough. In other words, transmission makes up for the smaller price differentials with increased volume of electricity transferred.
On the other hand, the value of transmission deriving from different prices between locales are assumed to be negligible in this example. A look at a map of average electricity prices in the United States
shows that most East-West transmission lines across time zones will also benefit from highly significant differences in average electricity prices. Since my example completely ignores these differences, my simple example is likely to greatly underestimate the value of transmission relative to storage.
What Does it Cost?
According to the Electricity Storage Association, the best storage technologies other than pumped hydropower (which I exclude from this analysis because new pumped hydropower developments are severely limited due to environmental and water rights issues) can store electricity for an incremental cost of about 2-5 cents per kilowatt-hour.
In contrast, Wikipedia puts the cost of "[l]ong-distance transmission of electricity (th
ousands of miles) [at] US$ 0.005 to 0.02 per kilowatt-hour." A one time-zone transmission line needs to be long enough to cover 15 degrees of longitude. This is slightly over 1000 miles at the equator, but gets smaller as you move to higher latitudes. In the United States, this varies between approximately 800 and 900 miles. The shorter distance should put the cost of transmission at the low end of the range above, or 0.5 cents per kWh, or about a quarter to a fifth the cost of storage.
In my simple example, the cost of transmission per kWh will have to be increased to compensate for the greater number of kWh transferred, since a storage system would likely only be cycled once per day (for two hours of charge or discharge), while the transmission system will be operating whenever the hour-to-hour price differentials are high enough to make up for line losses.
A look at hourly prices for electricity in Ontario for December 3rd (the day I am writing this) show that prices are rising or falling significantly about 16 hours a day, meaning we expect the transmission line to handle about 8 times as many kWh as the electricity storage. Hence the incremental costs for transmission may need to be increased from the above estimates to reflect the greater incremental costs due to line losses. However, all the "cents per kWh" numbers above contain assumptions about frequency of usage to allocate capital, maintenance, and electricity losses between kWh used. Storage technology for grid stabilization is most likely assumed to cycle approximately once per day, while this paper uses the assumption that a transmission line will be operated at 65% capacity, similar to the line in my example. Therefore, if the cost for transmission needs to be scaled up to reflect the higher usage, it should likely be increased by a factor significantly below 8, and perhaps not at all.
Taking this together, the value of the ability of the transmission line to act like electricity storage should be between 1 and 4 cents per kWh, still slightly below the 2 to 5 cents per kWh for most storage technologies. If we assume that there are any significant other benefits to transmission (such as increased diversification of power supplies and the ability to buy low and sell high between different regions, as discussed above), electricity transmission becomes the clear winner where it can be built.
Clearly, my comparison for a trans-time-zone transmission line and electricity storage is still far too simple to capture the full benefits of either technology. However, in an age when storage technologies are still mostly experimental while transmission technologies are well-established, it seems clear to me that our first efforts should be to capture those large-scale gains we can with a robust national grid. With President-Elect Obama promising "green" infrastructure spending to jumpstart the economy, neglecting electricity transmission would be a tragic mistake.
What’s Stopping Us?
To a believer in free markets, it’s probably quite surprising that such large economic opportunities exist. Similar to the untapped opportunities in energy efficiency, market barriers have crippled the national electric grid. The most obvious barriers are those of people who object to how they look. Because of the need for long, contiguous corridors, negotiation with individual landowners has delayed many projects for years. For many large projects, the power of eminent domain is essential. This is why T. Boone Pickens combined his wind plan with plans for a water pipeline from the Ogallala Aquifer. In Texas, water projects have eminent domain, while electricity projects do not.
There are also significant regulatory barriers. Electricity deregulation in many states meant that utilities often had no incentive to invest in new transmission infrastructure[pdf]. Furthermore, electricity planning is done state-by-state and region by region, with the North America carved up into nine independent regions.
Currently these problems are only being addressed on a state and regional basis. A robust national grid will require all these problems to be addressed at a national level. One approach might be for Congress to create a national transmission planning authority with the right of eminent domain, or the right to use the right of ways along the interstate highways system.
Is that too much to hope?