TODAY'S STUDY: THE BENEFITS OF WIND AND SOLAR TOGETHER

Transmission Benefits of Co-Locating Concentrating Solar Power and Wind

Ramteen Sioshansi and Paul Denholm, March 2012 (National Renewable Energy Laboratory/Ohio State University)

Executive Summary

This report discusses and analyzes the potential benefits of co-locating wind and concentrating solar power (CSP) plants in the southwestern United States. Using a location in western Texas as a case study, we demonstrate that such a deployment strategy can improve the capacity factor of the combined plant and the associated transmission investment. This is because of two synergies between wind and CSP. One is that real-time wind and solar resource availability tend to be slightly negatively correlated. The other is that low-cost and highly efficient thermal energy storage (TES) can be incorporated into CSP. TES allows solar generation to be shifted and used to fill-in excess transmission capacity not being used by wind. Adding TES in a transmission constrained system can reduce, but not eliminate curtailment, especially during periods of extended high wind output and high solar output. Adding transmission constraints associated with co-location also reduces performance, including the inability of CSP to provide maximum output during periods of both high demand and significant wind output. Overall the economic tradeoff between transmission costs and system performance is highly sensitive to assumptions regarding transmission and CSP costs. Using data from the years 2004 and 2005, we demonstrate that a number of deployment configurations, which include up to 56% CSP (on a capacity and energy basis) yield a positive net return on investment.

click to enlarge

Introduction

One of the challenges in deploying renewable electricity generation is the often remote location of high-quality wind and solar resources, requiring new transmission. The state of Texas was one of the first regions of the United States to contend with this issue. The highest quality wind resources are located in the western part of the state, while the major population centers are in the east. About 1.4 GW of wind was added in the McCamey region of Texas in 2001 and 2002, despite there only being about 400 MW-e of transmission capacity [1]. This resulted in about 380 GWh of wind generation being curtailed at an estimated cost of more than $21.4 million in 2002. Although substantial transmission capacity has been added, this construction has not kept pace with wind development, and nearly 8% of potential wind generation in the Electric Reliability Council of Texas (ERCOT) was curtailed in 2010 [2]. New transmission is often difficult to construct, and, if it is only carrying wind, these lines will tend to be lightly loaded due to the low capacity factor of the generator. For instance, a typical wind plant will have a capacity factor of less than 40%. One option to enhance utilization of new transmission is to co-locate wind with other resources that can complement their generation. This can include conventional generation, energy storage, or other renewables. Denholm and Sioshansi [3] demonstrate that energy storage is one viable option, since transmission can be downsized (relative to the capacity of the wind generator) and generation that would be curtailed can be stored for later use.

Another option is to deploy concentrating solar power (CSP) in areas with good wind resources. West Texas has good wind resources (at least class 4, which is defined as having average wind speeds of 7 m/s at a 50 m height) that are in close proximity to locations with sufficient direct normal irradiance (DNI) for economic siting of CSP plants (average daily DNI of at least 6 kWh/m2 is typically viewed as the minimum solar resource requirement). CSP plants have an additional advantage of being able to incorporate low-cost and high-efficiency thermal energy storage (TES), enabling them to become a dispatchable resource…

In this paper we examine the basic feasibility and performance of wind and CSP plants in western Texas that feed into the grid on a radial transmission line. We examine cases in which various combinations of wind and CSP are deployed with different amounts of transmission capacity. Using historical market price and weather data we model the operation of such a deployment and demonstrate how the dispatch, the capacity factor, energy curtailment, and the cost of energy are influenced by the combination of wind and CSP. The remainder of this paper is organized as follows: Section 2 provides further background on wind and CSP development, Section 3 discusses the methods and data used in the analysis, Section 4 summarizes our results, and Section 5 concludes.

click to enlarge

Dispatch of Wind and CSP Deployments

Wind and CSP have two basic complementarities that potentially enable them to share transmission. One is that real-time wind and solar resource are often negatively correlated. The other is that CSP with TES is partially dispatchable, which allows solar generation to be shifted and to fill-in excess transmission capacity during subsequent hours with lower wind and solar resource. This allows for increased transmission utilization…

There are, however, limitations to these complementarities between CSP and wind. Some periods, which tend to occur in the spring, will have extended high wind generation, during which the TES system of the CSP plants will be filled and solar thermal energy will have to be curtailed due to the limited transmission. Figure 7 demonstrates this for the same deployment during a two-day period beginning on April 22, 2005. Figure 7a assumes the base case of four hours of TES and in this two-day period about 71% of the energy collected by the CSP plants’ solar fields is curtailed. Overall, about 15% of the energy collected by the solar fields is curtailed annually, indicating that additional transmission or TES capacity may be desirable. Figure 7b shows the benefits of incorporating eight hours of TES, which somewhat reduces curtailment, and allows a more continuous output, for example largely filling the gap in hours 0 through 8 and 44 through 48. Figure 7a also shows that the model prioritizes wind over CSP. This is due both to the wind PTC, which makes wind $19/MWh-e more valuable than CSP from a marginal revenue perspective, and the higher variable operating and maintenance cost of CSP.

click to enlarge

Another limitation of having a downsized transmission link is that the CSP plant must be dispatched around wind generation and is operated in a suboptimal manner compared to a transmission-unconstrained deployment. This is illustrated in Figure 8, which shows the optimized dispatch of a deployment with 480 MW-e of CSP and 600 MW-e of wind during a three-day period beginning on October 17, 2004. For clarity, the dispatch of CSP is shown at the bottom of the figure. Figure 8a shows the operation of the deployment if there is no transmission constraint (i.e., there is 1,080 MW-e of transmission capacity). The figure shows that incorporating TES into the CSP plants allows the plant to sell energy during the highest-priced hours. For instance, stored thermal energy is discharged in hours 18, 30, and 62 when prices are relatively high…

Transmission Downsizing in Wind and CSP Deployments

Figures 6 and 7 point to the fundamental tradeoffs in deploying wind, CSP, and transmission. Reducing transmission capacity will increase the capacity factor of the transmission investment, but will also increase generator curtailment and decrease the value of the energy produced by the plant. Figures 8 through 12 explore these tradeoffs in greater depth. In addition to the base case with four hours of TES, the figures also include cases in which the CSP plants have eight hours of TES. Figure 9 shows the capacity factor of the transmission line as a function of the transmission capacity and generation mix of the deployment. The transmission and CSP capacities are given as a percentage of the 1,080 MW-e generating capacity of the deployment. The wind generators that we model have capacity factors over the two years studied of about 33.9%, whereas the CSP plants with an SM of 2.0 and four hours of TES have a capacity factor of roughly 34.5%. Thus if the deployment is transmission-unconstrained, a CSP-only generation mix maximizes the capacity factor of the transmission link. For a smaller transmission link, however, it is beneficial to build a mix of wind and CSP, in order to exploit the negative correlation between wind and solar resource. The mix that maximizes transmission capacity factor ranges between 44% (480 MW-e) CSP with a 600 MW-e transmission link up to 89% (960 MW-e) CSP with 1,000 MW-e of transmission capacity…

click to enlarge

Wind, on the other hand, tends to be slightly negatively correlated with energy prices. This negative correlation is shown in Figure 12 since the average selling price of energy from a wind-only deployment is very slightly decreasing in the transmission capacity. As the transmission capacity of the deployment is decreased, wind generation is curtailed in high-wind hours, which tend to be low-price hours. The other property of CSP is its dispatchability. This dispatchability allows CSP generation to be shifted, within the constraints of the TES system and transmission link, to higher-priced hours. This generation shifting is advantageous because, while prices and solar resource tend to be correlated, they are not perfectly coincident. This is because cooling loads can often lag the peak in DNI by a few hours due to thermal inertia. Thus, adding TES to a CSP plant allows for further energy revenue increases…

click to enlarge

Both Figures 11 and 12 show the effect of downsizing the transmission capacity, which is to diminish the value of TES. This is because CSP must share the scarce transmission with wind, as illustrated by the example in Figure 6. Thus, as the transmission capacity of the deployment is reduced, TES is decreasingly used to shift solar energy to high-price hours and is used instead to shift stored energy to low-wind and low-solar hours. Figure 13 also shows that increasing the amount of CSP in a deployment reduces the average selling price of CSP. This is, again, because the limited transmission capacity forces TES to be increasingly used to shift CSP generation to lower-price hours. As transmission constraints increase, the reduction in average price in the eight-hour TES case is due to the fact that while additional TES allows more energy to be sold, it tends to dispatch in lower-price hours. This means that the benefits of reduced curtailment associated with larger amounts of TES are somewhat offset by the lower value of this additional energy.

Figures 13 and 14 illustrate how this interaction between transmission and TES impacts deployment revenues. The figure shows average annual revenues as a function of the deployment configuration. Figure 14 shows total energy revenues from wind and CSP, excluding the wind PTC, and Figure 15 shows CSP energy revenues only. The figure shows that CSP earns considerably higher energy revenues than wind. This is both due to CSP having a slightly higher capacity factor (i.e., producing more energy) and because CSP produces higher-value energy due to the coincidence between DNI and prices and the use of TES. The figure also shows that downsizing transmission can significantly reduce deployment revenues. This has a greater effect on CSP—in the four-hour TES case, reducing the transmission capacity from 1,080 MW-e to 600 MW-e reduces annual CSP revenues by 26% (which amounts to $61 million for an all-CSP deployment), whereas it reduces wind revenues by 22% (which is $36 million for an all-wind deployment). The greater effect on CSP reflects the fact that much of the added value of CSP derives from its ability to use TES to deliver energy during high-price hours, which is limited when the transmission constraint is imposed. Figure 14 also shows the small total revenue difference between 4 and 8 hours of storage in CSP-only systems without transmission constraints.

click to enlarge

Long-Term Economics of Co-Located Wind and CSP Deployments

The analysis done thus far assumes the configuration of a deployment and analyzes short-term operational decisions. A related question is what deployment configuration would maximize long-term economic value. Such an analysis would require comparing upfront capital costs against discounted revenue streams over the lifetime of the different deployments. This type of an analysis presents four challenges. One is that we only have two years of time-coincident wind and solar resource data, which limits our ability to estimate revenues over the full lifetime of a deployment. A second is that the cost of utility-scale CSP is rather uncertain due to fluctuations in commodity prices and the potential for substantial manufacturing improvements. A third is that the cost competitiveness of wind and CSP relative to conventional generating technologies will depend on future policy decisions, such as carbon regulation and fuel prices and subsidies. Finally, this analysis does not consider the impact of the increased use of renewables on the ERCOT market, mix of generators and the value of dispatchable resources, such as CSP with TES…

click to enlarge

Revenue Estimates…Cost Estimates…Deployment Subsidies…Deployment Breakeven Costs…Deployment Investment Cost Recovery…

Conclusions

This report discusses and analyzes the potential benefits of co-locating wind and solar generation in the southwestern United States. Such a deployment strategy can improve the capacity factor of the combined plant and the associated transmission investment, with dispatchable CSP shifting energy to periods during reduced wind output. However, adding transmission constraints associated with co-location reduces performance, including the inability of CSP to provide maximum output during periods of both high demand and significant wind output. Furthermore, even with the use of TES, there are periods of extended high wind output and high solar output, which can result in curtailed energy. Despite these limitations, we do find cases in which a mix of CSP and wind can be justified by market revenues, and that in some cases deployments with up to 56% CSP capacity yield a positive net ROI. This will depend on a substantial reduction in CSP costs. Deployment economics are highly sensitive to transmission development costs, which have varied significantly in past projects, as well as future prices of conventional energy and potential restrictions on carbon emissions.

Tweet