5.2 Storage and capacity factor

A major storage advantage of a CST system over a PV system is that it can use thermal energy storage (TES) [29]. Storage for CST is more established than for PV with operational CST plants utilising TES throughout Spain and the USA. The authors are not aware of any large scale PV plants utilising storage, whereas SEGS I, which has three hours of TES, has been operating since 1985.

An example of how the power output from a solar plant (in this case CST trough plant combined with TES) can be shifted through utilising storage to more closely match a load profile is shown in Figure 52. A thermal storage system for CST can be as simple as:

  • a storage tank added to the loop through which the heat transfer fluid (HTF) flows
  • the flow rate between the collectors and the storage tank, and the storage tank and the heat exchanger, are controlled to ensure a constant supply of heat to the heat exchanger, giving constant power output from the steam turbine
  • the storage tank is drawn down during periods of cloud cover (or at nights) and re-filled during normal sunshine -smoothing out output variability.

Figure 52 The effect of storage [82]

A key advantage of TES in CST systems is that no additional energy conversion process is required resulting in more efficient addition of storage, unlike battery or mechanical storage for PV. Recent cost estimates puts the cost of adding TES to a CST system between $72 and $240 per kWh of electric storage capacity, along with high round trip efficiencies of up to 98% [29]. Table 7 gives estimates for the costs of battery storage using lead acid and VRLA types at $150 and $200 per kWh respectively, being the only types which would be considered competitive to TES [30].

Table 7 Batt ery cost ($/kWh) [30]

Technology Current Cost ($/kWh) 10-yr Projected Cost ($/kWh)
Flooded Lead-acid Batteries $150 $150
CRLA Batteries $200 $200
NiCd Batteries $600 $600
Ni-MH Batteries $800 $350
Li-ion Batteries $1300 $150
Na/S Batteries* $450 $250
Zebra Na/NiCl Batteries $800 $150
Vanadium Redox Batteries 20 kWh = $1800/kWh
100 kWh = $600/kWh
25 kWh = $1200/kWh
100 kWh = $500/kWh
Nn/Br Batteries* 30 kWh/45 kWh = $500/kWh
2 MWh = $300/kWh
Lead-carbon Asymmetric Capacitors (hybrid) $500 <$250
Low-speed Flywheels (steel) $380 $300
High-speed Flywheels (composite) $2500/kW $800
Electrochemical Capacitors $356/kW $250/kW

The increase in the Capacity Factor (CF), defined as the ratio of actual to potential power (nameplate size) supplied, of PV and CST plants with the addition of energy storage is also of interest. The 19.9 MW Torresol Gemasolar solar tower in Spain is able to manage 74% capacity factor with 15 hours of molten-salt storage [31]. Solar towers without storage can generally manage 20-25% capacity factor.

PV systems without storage seem to offer similar capacity factors as CST without storage [32]. This is possibly due to the reliance of CST systems on direct beam irradiance and without storage, the capacity factor depends on the relative overall solar to electric power conversion efficiency of the technology itself. Figure 53 shows capacity factors for a number of PV systems, ranging from 13.4% to 26.9% with an average of around 23.6%. The findings are based on 30 years of hourly irradiance data, each system being a singular module ranging from 50W to 250W. Equivalent capacity factors between CST and the PV systems can be seen, but it is worth noting that the PV systems are single modules. Capacity factors would likely be different if these systems were analysed in aggregate.

Figure 53 PV capacity factors [32]

Research in [33] considered how battery storage helped decrease the Unserved Critical Load (UCL) in a high PV penetration scenario. UCL is a proxy for SAIFI3, which is defined as the annual average number of critical load interruptions experienced on a circuit. The study used ten, a hundred and a thousand houses in three different areas of the US, and compared UCL levels for systems with no storage and 1 kWh worth of battery storage. Figure 54 shows the results of the study. The most significant result was for the California (CA) 1000 houses case, where the UCL dropped from around 85% to 50%. This can alternatively be seen as a 35% increase of critical load served - this scenario consists of 1000, 2.1 kW PV systems with 1 kWh battery storage.

Figure 54 Comparison change in UCL due to battery storage [33]

This comparison between CST and PV systems looking into how much storage improves capacity factor is not an ideal one as it should really be done between systems of equivalent generation and storage capacity, and a large number of systems should be compared. However, there do not appear to be any large scale PV plants utilising storage. Note that the PV scenario used for the comparison is the best of those presented in Figure 54, and the Torresol Gemasolar solar tower has the best capacity factor of all operational CST systems. Although the comparison is somewhat mismatched, it gives some indication of the effectiveness of storage in improving the capacity factor for both CST and PV systems.