Concentrating solar power (CSP) systems which use concentrated sunlight to run steam turbines have been receiving a lot of attention in recent years as a potential low cost alternative to photovoltaic cells. Like all solar technologies the power delivery profile of CSP depends on the availability of sunlight. Adding energy storage to such systems increases their power delivery flexibility. Unlike PV cells CSP systems can potentially store thermal energy rather than electrical energy giving them a cost advantage in this area because sensible heat storage has lower cost than electrical storage in batteries.
Sensible heat storage for a power tower CSP system was orginally demonstrated in a project funded by the U.S. Department of Energy called Solar Two. This project used a mixture of nitrate salts as the thermal storage medium. Some details of this project taken from the paper Solar Two: A Molten Salt Power Tower Demonstration by Craid E. Tyner, J. Paul Sutherland, and William R. Gould are given below.
In parallel with Solar One, a series of studies funded by the U.S. Department of Energy and industry examined advanced power tower concepts using single-phase receiver fluids, the best of which was a 60% sodium nitrate/40% potassium nitrate molten salt. The primary advantages of molten nitrate salt as the heat transfer fluid for a solar power tower plant include a lower operating pressure and better heat transfer (and thus higher allowable incident flux) than a water/steam receiver. This translates into a smaller, more efficient, and lower cost receiver and support tower. In addition, the relatively inexpensive salt can be stored in large tanks at atmospheric pressure, allowing 1) economic and efficient storage of thermal power collected early in the day for use during peak demand periods; 2) increased plant capacity factor by oversizing of the collector and receiver systems with storage of the excess thermal energy for electricity generation in the evening; 3) isolation of the turbine-generator from solar energy transients; and 4) operation of the turbine at maximum efficiency. If necessary, a molten salt system can be hybridized with fossil fuel in a number of possible configurations to meet demand requirements when the sun is not shining.
A schematic of a molten salt power tower system is shown in Figure 2. During operation, cold (285°C) molten salt is pumped from the cold salt tank through the receiver, where it is heated to 565°C. It then flows by gravity to the hot salt tank, where it is stored until needed for generation of steam to power the turbine. At that time, it is pumped through the steam generator, producing 512°C steam for the electric power generation system before being routed back to the cold tank to begin the cycle again.
A 43-MWth external, cylindrical nitrate salt receiver replaces the Solar One water/steam receiver. The Solar Two receiver is 5.1 m in diameter, 6.2 m tall, and receives an average flux over 0.4 MWth/m2 from the heliostat field. It uses 24 panels of 32 tubes each, terminated in headers at each end. The tubes, made of 316 stainless steel, have an inside diameter of 2.06 cm and a wall thickness of 0.12 cm. Molten salt enters the receiver at 285°C, flows in a serpentine path in two parallel control zones through the receiver, and exits the receiver at 565°C. The receiver was designed and manufactured by Rockwell International. Figure 5 shows the installation of the new Solar Two receiver. A nitrate salt thermal storage system has replaced Solar One's oil/rock thermocline unit. Sized for 3 hours of full turbine output, the storage system uses separate cold (285°C) and hot (565°C) salt tanks. The system will demonstrate the decoupling of solar energy collection from electric energy generation, the potential to meet a utility evening peak demand, and the dispatch characteristics of a commercial plant. The tanks, designed and erected by Pitt-Des Moines, Inc., are externally insulated and utilize air to passively cool the foundations to meet soil load-bearing constraints. The cold tank is 11.6 m in diameter, 7.8 m tall and is constructed of carbon steel; the hot tank is 11.6 m inside diameter, 8.4 m tall, and is constructed of stainless steel.
Approximately 1600 tonnes of salt are used in the system. Salt flows from the cold and hot nitrate salt storage tanks by gravity to two sump vessels (receiver sump and steam generator sump, respectively) in which the main nitrate salt pumps are mounted. The receiver sump is a 4.3-m diameter, 2.9-m tall hemispherical-head vessel. The steam generator sump is a 4.3-m diameter, 2.4-m tall flat-head vessel. The sump vessels also serve as low points in the system to drain salt from all associated piping systems. Cold salt is pumped from the receiver sump to the receiver by two multistage, vertical turbine pumps. Hot salt is pumped from the steam generator sump to the steam generator by two vertical cantilever pumps. The receiver pumps are 50% rated capacity with both pumps required to operate the receiver. The two steam generator pumps are 100% rated due to the limited size ranges available for the required flow, temperature, and pressure conditions, and to provide redundancy. (The project design approach has been to use commercially available equipment to the maximum extent possible and avoid developing special equipment.) In addition to the main pumps, a nitrate salt
mixer pump uses cold salt to attemperate the hot salt streams to the steam generator
and for the production of auxiliary steam.
An inherent design feature of nitrate salt technology is the electric heat tracing required for piping and components to prevent salt freezing (at approximately 240°C) in any pipe or component. The process electric heat tracing (EHT) system was designed and manufactured by Raychem. The primary function of the EHT is to warm-up and control the nitrate salt piping systems and equipment to prevent thermal shock and to maintain filled systems above the salt freezing temperature.
Thermal Storage: Externally insulated hot and cold salt tanks, 900 m3 each, 105 MWth (3 hr)
Turbine: Non-reheat Rankine cycle 10 MW net electric
Unlike the 100-MW plant, Solar Two does not use a reheat turbine cycle.
Consequently, gross Rankine-cycle efficiency will be reduced from 42.5% to 33%.
From the information given in the above extracts one can calculate an storage energy density in Whe/kg. Because the steam turbine was not optimized (see the last extract above) the effective electrical energy storage acheived was less than optimum. I correct for this effect by multiplying by a factor of the optimum Rankine-cycle efficiency/Achieved efficiency
Energy Density = (42.5/33)*(3hours*10E6 W)/(1600 tonnes*2000kg/tonne) = 12Wh/kg
The energy density is not particulary impressive compared to electrochemical batteries (The energy density of lead acid batteris is 25Wh/Kg), but the real issue is cost. Postassium nitrate and sodium nitrate are cheap, plentiful materials, and the salt storage system is likely to have a very long lifetime compared to electrical batteries which degrade relatively quickly with use. However, the comparative low energy density means that storage periods longer than a day or two are unlikely; Molten salt thermal storage is not a means to use summer sunlight to generate electricity in the winter.
The molten salt receiver for the Solar 2 project was developed by Rocketdyne Aerospace which, at that time, was a subdivison of Rockwell International. Since then Rocketdyne has been acquired by United Technologies, which has spun off the salt storage/power tower technology to a company called SolarReserve which is moving forward with plans for commercialization.