OPERATIONAL VALUE

The operational value of each technology represents its ability to avoid the variable cost of operation. These costs were tracked in three cost catego- ries—operating fuel, variable O&M, and start-up costs. Operating fuel in­cludes all fuel used to operate the power plant fleet while generating and includes the impact of variable heat rates and operating plants at part load to provide ancillary services. Start-up costs include both the start fuel, as well as additional O&M required during the plant start process. In each case the operational value was calculated by dividing the total avoided generation cost in each cost category by the total potential solar generation. Table 4 summarizes the results from the production simulations.

TABLE 4: Operational Value of Simulated Generators

Marginal Value ($/MWh)

Low RE

High RE

Flat

Block

PV

CSP

(no TES)

CSP (6-hr TES)

Flat

Block

PV

CSP

(no TES)

CSP (6- hr TES)

Fuel

31.7

35.2

33.9

37.7

22.6

21.2

18.7

31.1

Variable

1.2

1.0

1.0

0.8

2.1

2.0

1.9

1.4

O&M

Start

0.4

0.4

0.6

3.5

0.5

-0.9

-1.7

3.1

Total

33.3

36.6

35.5

42.1

25.2

22.3

18.9

35.6

Table 4 demonstrates three significant findings: (1) at low penetration, the value of solar generation technologies is greater than the constant (flat block) resource; (2) the value of all generation decreases as a function of renewable penetration, but the value of non-dispatchable solar resources decreases at a greater rate than the flat block or dispatchable CSP; (3) the value of CSP with storage is higher than solar technologies without stor­age. The range of values for different generation technologies largely can be explained by understanding the avoided fuel mix in the two different scenarios.

In the test system, the added generators (flat block or solar) reduce the output from a mix of generator types and with different efficiencies, depending on the time of day and season. Figures 9-16 illustrate how the relative value of a renewable generator is affected by the varying marginal generators and the dispatchability of the resource.

Figure 9 illustrates the relationship between price and net load for a 3-day period starting on January 22. The net load is the normal load minus wind and solar PV generation and reflects the load that must be met by other (mostly fossil fueled) generators with non-zero generation cost. The figure illustrates three zones of prices, which are seen earlier in the price

image183

FIGURE 9: System net load and marginal price for January 22-24 (low RE case)

duration curves in Figure 8. The lowest price occurs in the overnight peri­ods at the beginning of days 2 and 3 when coal is the marginal generator with total incremental cost of about $20/MWh. During much of the middle of the day, combined cycle units are the marginal generators, with variable costs of about $30-$35/MWh. In several periods in the morning and eve­ning, there is an increase in net demand, where the high ramp rate or the relatively short period of increased demand requires the use of combustion turbines, resulting in a price spike to about $45/MWh. Any renewable gen­erator added to this mix will offset energy within these three price zones but with a value depending on the temporal pattern of its output.

Figure 10 keeps the marginal price curve but adds the generation pro­file for CSP with and without storage. The CSP dispatch is isolated from cases where CSP is added. The total generation by these two plants is very similar, but CSP with storage is dispatched during the highest cost periods. In much of the winter, the price of electricity peaks in periods where solar output is low or zero (the morning and evening). This cor­responds to when higher-cost gas-fired units are started and ramped to meet peak demand. PV and CSP without storage are unable to gener­ate during this period and typically offset more efficient gas-fired units.

image184

image185

FIGURE 11. System. price for July 14-16 (low RE case)

Alternatively, CSP with TES is able to shift generation to the evening and carry over energy to start and pick up the morning load ramp that occurs before significant solar energy is available. As a result, CSP avoids the use of higher-cost and lower-efficiency gas-fired units, producing overall higher value to the system.

During the summer, operation of CSP with storage is more continuous due to higher solar output and a different load and price profile. Figure 11 shows the relationship between net load and system marginal price for a 3-day period starting on July 14.

The corresponding CSP operation is shown in Figure 12. There are several operational issues that affect the overall and relative value of CSP with TES. First, CSP with TES is able to operate more continuously and avoid the impact of cloud cover that reduces output and increases the vari­ability of the plant without TES. Second, CSP is able to start earlier in the day and help pick up the early morning load ramp. Finally, CSP is able to continue operation longer into the late afternoon and early evening. This is particularly important for the plant capacity value discussed in Section 5.3.

image186

FIGURE 12: System marginal price and corresponding CSP generation on July 14-16 (low RE case)

Figure 12 shows the impact of the solar multiple, which can provide some disadvantages at low solar penetration. At low penetration of solar, PV and CSP without storage are largely coincident with demand (and rela­tively high prices) during the summer. As illustrated in Figure 2, whenever the thermal output of the solar field exceeds the power block capacity, en­ergy must be stored, regardless of the system demand for energy or price. As a result, the plant is forced to store this energy and generate at a later time, even if this later time has a lower demand or lower cost of energy. This is shown in Figure 12 on the first and second day, when during some hours, CSP without storage sells more energy at periods of high prices than CSP with TES. CSP with storage is forced to shift some energy to the evening when prices are slightly lower.

The value of solar and dispatchable CSP is strongly dependent on the mix of generator types and amount of renewable energy. As the penetra­tion of renewables increases, the patterns of net demand for electricity change, and different mixes of generation are needed to address the in­creasing variability and uncertainty of the wind and solar supply. Figure

13 is a duplication of Figure 9, showing price and load during 3 days in January, except for the high RE case. The large amounts of wind and solar PV have suppressed the marginal price, and coal is on the margin for more hours. The load shape (and price) is also much more volatile, with opera­tion of combustion turbines to address the shorter peaks.

In the high RE scenario the absolute value of all energy sources drops due to lower system marginal prices. However, the value of variable en­ergy sources drops at a much faster rate than dispatchable sources, as a plant with TES is able to change output to capture the remaining periods of high prices. Figure 14 shows how CSP with TES is able to generate during the hours of highest price during this period in January.

Dispatchability becomes increasingly important during periods of very high renewable output to avoid generating during periods of zero value and associated renewable curtailment. Figure 15 shows a period of low net demand due to high solar (and wind) output during the middle of the day. During the first 2 days shown shortly after noon, the net load drops to the point where all thermal generators have reduced output to their minimum.

image187

FIGURE 13: System net load and marginal price for January 22-24 (high RE case)

image188

FIGURE 14: System marginal price and corresponding CSP generation on January 22-24 (high RE case)

image189

FIGURE 15: Net load and price for a 3-day period starting February 8

Подпись: 266 Solar Energy: Application, Economics, and Public Perceptio

image191

FIGURE 16: System marginal price and corresponding CSP generation on February 8-10 (high RE case)

 

This is the same period as the first 3 days in Figure 8, where during the middle of the day all coal plants in the system cannot reduce output further without incurring a costly shut down. Any additional zero-cost renewable energy generated during these hours cannot be used by the system so have zero value, and the system marginal price is $0/MWh.

Figure 16 shows the operation of CSP plants in these three days. CSP without storage generates in the middle of the day, producing some output that provides zero incremental value (when the system marginal price is zero). During these periods, CSP with thermal storage generates at low output, or shuts down, avoiding curtailed energy and maximizing value by shifting energy to periods of higher net demand and providing potentially valuable ramping services.

The ability to avoid renewable generation during periods of low or zero value will be an increasingly important source of value as renewable pen­etration increases. In the high RE scenario, about 5% of the additional PV and about 6% of the CSP without storage has zero value and is effectively curtailed. The number of hours of zero value generation (resulting in re­newable curtailment) is highly non-linear as a function of renewable pen­etration and would be expected to increase without additional measures to increase system flexibility (Denholm and Margolis 2007).

The sum of these factors, including the mix of generation, fuel cost, and curtailment can be translated into the source of avoided fuel costs in Table 4. Tables 5 and 6 further explain the source of avoided costs for the different generator types. Table 5 indicates the type of generation avoided by each unit of generation. In the low RE case, each kilowatt-hour of CSP without storage avoids 0.9 kWh of combined cycle generation and 0.1 kWh of combustion turbine generation. In some cases, the smoothing of the load can actually increase the use of some lower-cost generator types; in the low RE case, CSP with storage and PV can improve the system dis­patch and increase low-cost coal generation slightly. The flat block results show a greater displacement of coal because it generates at constant out­put, including at night when coal is often on the margin. Table 5 demon­strates how, in the high RE scenario, much of the gas generation has been removed by the system, and coal is on the margin for more hours. Both PV and CSP without storage remove similar amounts of combined cycle and coal; however, CSP with storage continues to avoid mostly gas generation due to the dispatchability of the resource.

In some cases, each unit of generation removes more or less than 1 unit of thermal generation. This is due to two factors: pumped storage operation and curtailment. The flat block in the low RE case frees up coal generation to displace more costly gas plant operation via the use of pumped storage. However, because storage incurs losses, this results in a small increase in thermal generation. The opposite occurs in the low RE PV case and the CSP with storage cases. The displacement of higher-cost generation in these cases reduces the economic operation of pumped storage, decreas­ing storage losses and resulting in more than 1 unit of avoided generation per unit of solar generation. At higher RE penetration, solar without stor­age displaces less than 1 unit of thermal generation due to curtailment of renewable generators.

TABLE 5: Avoided Thermal Generation

Avoided Thermal Generation (kWh/kWh)

Low RE Scenario High RE Scenario

Flat

Block

PV

CSP

(no TES)

CSP (6- hr TES)

Flat

Block

PV

CSP

(no TES)

CSP (6- hr TES)

Coal

0.09

-0.06

-0.03

-0.08

0.55

0.50

0.52

0.17

Gas Combined

0.78

0.99

0.91

0.79

0.39

0.50

0.39

0.72

Cycle

Gas Turbine/

0.10

0.09

0.11

0.27

0.05

-0.02

-0.01

0.11

Steam

Total

0.98

1.03

1.00

1.02

0.99

0.97

0.90

1.04

While Table 5 is a useful illustration of the type of generation avoided, the ultimate cost driver is the type and amount of fuel actually displaced. Table 6 provides the actual avoided operational fuel in each scenario (in MMBTU per MWh of solar generation). Of note is the fact that the avoid­ed fuel rate increases in the high RE scenario. This is due to the displace­ment of lower cost, higher heat rate coal units compared to more efficient, higher-cost gas generators. The product of the avoided fuel in Table 6 and fuel costs produce the fuel value ($/MWh) in Table 4.

TABLE 6: Avoided Fuel

Avoided Fuel (MMBTU/MWh)

Low RE Scenario High RE Scenario

Flat

PV

CSP

CSP (6-

Flat

PV

CSP

CSP (6-

Block

(no TES)

hr TES)

Block

(no TES)

hr TES)

Coal

1.1

-0.7

-0.7

-0.9

5.8

5.2

5.4

1.9

Gas

7.4

8.9

8.9

9.7

3.5

3.6

2.9

7.1

Total

8.5

8.2

8.2

00

00

9.3

00

00

8.3

9.0

An additional important secondary source of value for CSP with TES is the ability to avoid thermal plant starts and associated fuel use and main­tenance. Even at low penetration, PV and CSP without storage tends to in­crease the variability of the net load, increasing the number of plant starts but decreasing the total amount of energy produced by the generation fleet. Table 7 provides the estimated number of avoided starts and percentage re­duction. Consistent with the previous tables, a positive number represents actual avoided starts (a net benefit), while a negative number means an increase in starts. This table demonstrates a significant reduction in starts due to the flexible operation of CSP with TES.

TABLE 7: Avoided Starts

Avoided Starts (Total/%)

Low RE Scenario

High RE Scenario

Flat

PV

CSP

CSP (6-

Flat

PV

CSP

CSP (6-

Block

(no TES)

hr TES)

Block

(no TES)

hr TES)

Coal

3/

-1/

-5/-0.7%

-8/-1.1%

-16/

4/

-3/

-18/

0.4%

-1.1%

-2.2%

0.6%

-0.4%

-2.5%

Combined

-77/

18/

53/4.3%

56/4.6%

9/1.2%

17/

-56/

129/

Cycle

-6.3%

1.5%

2.2%

-7.2%

16.5%

Gas

362/

-412/

-271

1,099/

432/

-640/

-361/

871/

Turbine/

Steam

4.6%

-5.2%

/-3.4%

13.8%

4.1%

-6.1%

-3.4%

8.3%

image192

FIGURE 17: Correlation of demand and solar generation on a 3-day period starting July 26 (low RE case)