3.2. The role of energy storage systems in hourly and seasonal operations
To demonstrate the changing role of energy storage systems over the time frame analyzed, let's first look at other drivers of power system change. For example, the rapid deployment of solar power and wind power facilities has been an important factor in the changing power generation technology in recent years. Figure 6 shows the diurnal (daily) average net load (electricity demand minus generation) for a high gas cost/low cost battery scenario to 2050. In 2020, it is common to see a typical load curve that is lowest at night and highest in the evening in summer. By 2030, the massive increase in solar power generation (14% of total power generation) has changed the shape of the net load curve, reducing net load at noon and delaying evening peaks in the early evening. All seasons show a second morning peak before sunrise. These characteristics will persist and intensify in 2040 and 2050, as solar power generation will increase to 20% and 23%, respectively.
As the net load varies from year to year, so does the operation of the energy storage system (as shown in Figure 7). On the top panel showing 2020, the energy storage system is mostly discharged at night, which corresponds to the lowest net load level in Figure 7 and therefore the lowest marginal cost of generation. In all seasons in 2020, the discharge of the energy storage system is mainly carried out during the peak net load in the afternoon and evening. However, by 2030, energy storage systems have shifted to almost full charging at noon, in response to the overproduction of solar power generation facilities. While energy storage systems discharge at night, in 2030 they are typically discharged later in the evening as solar generation shifts peak periods to later in the day. It is also starting to see some power generation facilities supplying electricity during the morning rush hours. The same pattern continues and intensifies in 2040 and 2050, with energy storage systems charging at noon during the day and discharging before sunrise and after sunset that day.
There is little difference in the average daytime energy storage system usage between the four seasons of the year, especially in the non-summer months. This observation is generally true in Figure 8, which illustrates the percentage of total energy storage capacity used on a daily basis for all five scenarios for each month of the year. However, some interesting trends have been observed. First, in most non-summer months in 2030, the average usage percentage for some scenarios (zero carbon and low-cost PV) is slightly above 100%. In these cases, the shorter duration (2 and 4 hours) battery storage systems are charged at night (via wind or nuclear/coal, depending on the region) and discharged during the morning peak, then again during the day Use solar power to charge and discharge during evening rush hours. Figure 9 illustrates this with an example of five days in January 2030 under a low-cost solar power scenario. The top panel shows the daily power generation of the 2-hour battery energy storage system, usually showing two charge and discharge processes per day. Longer duration pumped hydro power generation facilities and 6-hour battery storage systems usually show a charge-discharge process per day. However, most years and scenarios show that charging and discharging for more than a day is still rare.
The report also notes that in 2030, the summer months will have the lowest utilization of energy storage systems. Net load rarely has a "duck curve" (net load remains high throughout the day) as additional cooling loads drive higher overall demand. Noon and afternoon loads these days are largely in line with the availability of solar power. Figure 10 illustrates this with 5 days in July 2030 for a low-cost solar power generation scenario. In summer, only one daily charging cycle is usually noticed, even for configurations of energy storage systems of short duration. It is also noted that there is less demand for energy storage systems in the morning (relative to winter), and almost all energy storage discharge operations take place in the evening peak.
In 2040 and 2050, due to the increase in storage deployment and very little seasonal variation, the utilization of energy storage systems will generally decrease, except for the early summer months of certain seasons. Energy storage system usage may increase . The early summer months typically showcase strong wind and solar generation, as well as milder temperatures than the late summer months bring moderate loads.
Overall, the overview of battery energy storage system operation in this section shows that energy storage system operation is closely related to solar generation availability and less so to wind generation. This conclusion is consistent across scenarios, asset mixes, and seasons. Solar power facilities have predictable daily on and off cycles, which are well aligned with the need to charge and discharge energy storage systems. On the other hand, wind power facilities often experience prolonged periods of overproduction, lasting hours or days, much longer than the energy storage systems studied here. While energy storage systems can play a key role in storing solar and wind power, synergies with solar power are more consistent.
3.3. Sensitivity of energy storage system
So far, the role of energy storage systems on future power systems on a daily, seasonal, and annual basis has been discussed. However, meaningful insights can also be gained by evaluating the role of energy storage systems. So far, the optimal installed capacity of energy storage through cost-optimized deployment in District Energy Deployment Systems (ReEDS) has been considered. For a "sensitivity" described in this section is an effect that the study reports on adjusting the installed capacity of the energy storage system in the high gas cost/low cost battery scenario to 2050 to observe its effect on the energy storage system. Value impact, interaction of transmission system and energy storage, and impact on traditional generator set dispatch.
For the sensitivities described in this section, a 660GW energy storage system deployed in a high gas cost/low cost battery scenario to 2050 is used as a benchmark. The installed capacity of the energy storage system is adjusted by ±5% "increments" of 80% to 120% of the energy storage system. Each increment will reduce or increase the total installed capacity of the energy storage system by about 33GW. It is emphasized here that the reduction and addition of energy storage from the generation mix deployed by the District Energy Deployment System (ReEDS) results in an expansion that is no longer “optimal”. This means that with less than 100% installed capacity of the energy storage system, reliability may be reduced. Although not seeing a large amount of load drop in one scenario, the low rate of energy storage system deployment did increase the incidence of energy storage system capacity drop relative to 100% energy storage susceptibility.
3.3.1. Incremental operational value of energy storage systems
Sensitivity (80%, 100, and 120%) power system dispatch for three energy storage systems over four days in May 2050. The 80% energy storage susceptibility indicates significant cuts (especially around noon on May 6, when cuts of more than 150GW were taking place). Both 100% energy storage susceptibility and 120% energy storage susceptibility cutbacks are declining, and are just below 100GW in 120% energy storage susceptibility. The daily dispatch also shows that in the 100% and 120% storage sensitivity scenarios, the solar generation time increases and the in-house natural gas generation decreases, relative to the 80% storage sensitivity.
Energy storage systems can efficiently use electricity with lower marginal costs (such as electricity from solar power generation facilities) and can replace more expensive electricity (such as electricity from natural gas combined cycle power generation facilities). Therefore, higher energy storage susceptibility shows lower annual cost of electricity generation.
It should be noted that the annual power generation cost increases as the energy storage system becomes less sensitive at 100% energy storage system, and decreases as the energy storage system is added. This curve illustrates how the incremental energy storage capacity will decrease as more energy storage systems are added to the power system. For example, between 80% and 85% storage susceptibility, adding 33 GW of storage could reduce annual generation costs by $732 million. In contrast, adding 33 GW of energy storage in the 115% to 120% energy storage scenario reduces the total cost by only $248 million. This study shows that as installed capacity increases, the arbitrage opportunities for energy storage systems will decrease.
Total annual cost of electricity generation for storage sensitivity (in billions of dollars). 2050 high gas cost/low cost battery scenario shown in the middle as 100% energy storage system sensitivity
Much of the value of an energy storage system comes from reduced fuel costs. It should be noted that the increment of the energy storage system is only a small part of the total value of the increment of the energy storage system. As mentioned above, energy storage systems also have capacity value. While the value of capacity can be represented by production cost modeling (such as energy storage system availability during the first 10 net load hours in Figure 3), the value of capacity and arbitrage will ultimately be related to costs in capacity expansion models such as ReEDS weighted.
