As more variable generation resources are added to the electric supply mix, especially wind and solar energy fueled generation, the electric supply will vary along with demand. That phenomenon along with the corresponding need for frequency regulation is shown graphically in Figure 5.
Historically, generation has provided most of the area regulation service. Generation provides up or down regulation exclusively, or it can be used to provide some of each, as shown graphically in Figure 6 below.
Similar to load following, storage provides area regulation up with increased discharge and/or reduced charging while it provides area regulation down via reduced discharging and/or increased charging. When storage charging is used to provide area regulation, storage related energy losses result in a real-time purchase of make-up energy. So storage used that way must have high efficiency (i.e., >90%).
Storage has important advantages. If the storage used is very efficient (i.e., charging can be used for regulation as well as discharging) then it can provide area regulation equal to two times its power rating, as depicted graphically, by the rightmost bar in Figure 6. That is because storage can provide both regulation up and regulation down both by discharging and by discharging, like load following, but faster.
Fast Frequency Regulation
Storage with a fast ramp rate and that can be configured to have 15 to 30 minutes of storage discharge duration is well-suited to provide area regulation. In fact, there are indications that storage with a high ramp rate is perhaps twice as valuable (i.e., effective) as generation-based area regulation. That is because most types of generation have a slow ramp rate, meaning that their output cannot be changed quickly.
Storage with a very fast ramp rate can provide the relatively new ancillary service called frequency response. Storage used for frequency response actually monitors the AC frequency and responds to anomalies, over time frames of milliseconds. The objective is to keep the frequency as close to the target frequency – 60 cycles per second in the United States – as possible. The concept of frequency and specifically 60 cycles per second AC frequency used in the United States is shown in Figure 7.Frequency response is similar to area regulation with an important distinction: Frequency response resources monitor the AC frequency and they respond to frequency excursions whereas area regulation responds indirectly based on control signals that reflect a difference between electric supply (power) electric demand (power). Also, output from frequency response resources changes much faster – in less than a second – than output from area regulation which changes every few seconds or minutes.
Currently there are few existing/conventional electric supply resources whose ramp rate is fast enough to respond to sub-second signals so fast storage is especially well-suited to this application.
Storage used for frequency response service should reduce the need for fast-responding generation that would otherwise be needed for area regulation and could reduce generation start-ups, output variability and part load operation which, in turn, reduce fuel use and air emissions.
Ramping is a significant change of generation power output over time frames ranging from a few seconds to a few minutes. Of particular interest are: a) wind generation ramping that is caused by rapid wind-speed variations and b) solar generation ramping which occurs as large clouds pass over the generator. Indeed, ramping may increase as variable resources are added to the grid. If ramping does become significant, then system operators will have to respond, or the grid could become unstable.
Similar to load following and area regulation, the ramping ancillary service involves resources that offset output ramping. So resources used for the ramping service provide output variability that is the reverse of other generations’ output variability due to ramping. Perhaps the best example is wind generation whose output ramps up or down quickly as wind speed changes quickly. In that case output from resources providing ramping service must increase or decrease commensurate with wind generation output changes.
The conventional ramping service resource is new and/or existing generation; however, as with area regulation, most generation is not very suitable for ramping because it must be capable of relatively rapid output changes.
For systems needing additional generation capacity, operators might install and operate additional combustion turbines. For grid systems with excess generation capacity, ramping may be accommodated using existing generation whose output can be varied rapidly.
Storage used for ramping service (in lieu of generation) provides ramping up by increasing output and/or by decreasing charging. Conversely, storage provides ramping down by decreasing output and/or increasing charging, as shown in Figure 8.
The benefits for use of storage for ramping service are:
- Reduced need for generation capacity,
- Reduced generation start-ups,
- Reduced generation output variability and “part load” operation (i.e., operation at below the generation’s maximum output), and
- A commensurate reduction of fuel use and air emissions.
Reserve capacity is essentially backup generation for the electricity grid, for use if one or two large power sources become unavailable unexpectedly. So, when using storage as electric supply reserve capacity, the need and cost for generation-based reserves is offset and, to a lesser extent, operation cost incurred for generation-based reserve capacity are reduced/avoided.
There are three generic types of reserve capacity:
- Spinning Reserve – Generation capacity that is online but unloaded and that can respond within 10 minutes to compensate for generation or transmission outages. ‘Frequency-responsive’ spinning reserve responds within 10 seconds to maintain system frequency. Spinning reserves are the first type used when a shortfall occurs. Also known as synchronized reserves.
- Supplemental Reserve – Generation capacity that may be off-line, or that comprises a block of curtailable and/or interruptible loads, and that can be available within 10 minutes. Unlike spinning reserve capacity, supplemental reserve capacity is not synchronized with the grid (frequency). Supplemental reserves are used after all spinning reserves are online.
- Backup Supply – Generation that can pick up load within one hour. Its role is, essentially, a backup for spinning and supplemental reserves. Backup supply may also be used as backup for commercial energy sales.
The amount of reserve capacity needed is driven by electric supply reliability-related standards (typically, 10 to 20% of the normal electric supply capacity). Given the focus on reliability, in the United States, the National Electric Reliability Council (NERC) is a key agency involved in establishing reserve capacity requirements.
When charged, storage can, in most cases, provide reserves merely by being ready to discharge. Furthermore, reserves are needed infrequently so storage used for reserve capacity is actually discharged infrequently. That gives storage an advantage over generation-based spinning reserves because generation used must actually be “spinning” and ready to pick up load on a moment’s notice.
Another attractive feature of storage-based reserves is that, while charging, storage can provide two times its capacity as reserves: Storage charging is stopped and the storage then the storage can be discharged.
The benefit from storage used for electric supply reserve capacity is somewhat small because generation-based reserves are inexpensive. Nonetheless, the reserve capacity benefit could be an important element of an attractive storage value proposition because providing reserves using storage has very low incremental cost.
An important technical challenge for electric grid system operators is to maintain the necessary voltage level and stability. In most cases, meeting that challenge requires management of reactance. To manage reactance at the grid system level, grid system operators rely on an ancillary service called voltage support.
Historically, voltage support has been provided by generation resources that can generate reactive power which offsets reactance in the grid. New technologies (e.g., modular energy storage, modular generation, power electronics and communications and control systems) make new alternatives for voltage support increasingly viable.
This is an application for which “distributed” storage (storage located close or very close to electricity end-users) may be especially attractive because reactive power cannot be transmitted efficaciously over long distances. Notably, many major power outages are at least partially attributable to problems related to transmitting reactive power to load centers. So, distributed storage – located within the load centers where most reactance occurs – provides especially helpful voltage support.
Black start resources are the first to power up to re-energize the grid after a grid-wide outage. Importantly, black start resources must be able to startup without power from the grid and must be able to operate in standby mode, while disconnected from the grid, until they are called upon. In most cases, the black start service is provided by specially-equipped generators. Most storage types are well-suited to serve as black start resources because, unlike generators, they do not need special equipment, and storage does not have to operate while awaiting dispatch.
Conclusions and Observations
Energy storage is especially well-suited to provide all necessary ancillary services needed for an efficient, stable and reliable electricity grid. Using storage for ancillary services reduces the need and cost for generation capacity and operation, reduces fuel use, reduces air emissions and frees-up existing generation to do generation is best at – generating electric power and energy.If the storage is distributed, it can provide other benefits and/or it can provide superior ancillary services.