862

Transactions of the American Nuclear Society, Vol. 116, San Francisco, California, June 11–15, 2017

Thermal Energy Storage Systems and their Integration with NPPs

Heat Storage for Peak Power with Base-Load Rankine-Cycle LWRs

�nd Brayton-Cycle High-Temperature Reactors

Charles Forsberg

Massachusetts Institute of Technology, 77 Massachusetts Ave., Cambridge, Massachusetts, 02139, U.S.A., cforsber@mit.edu

INTRODUCTION

The world is transitioning to a low-carbon energy system. Variable electricity and industrial energy demands have been met with storable fossil fuels—systems with low-capital costs and high-operating costs. In contrast, the low-carbon energy sources (nuclear, wind and solar) are characterized by high-capital-costs and low-operating costs. High utilization is required to produce economic energy.

The addition of large quantities of wind or solar changes the market. The Massachusetts Institute of Technology (MIT) Future of Solar Energy [1] study provides an examination of the solar option and the challenge of moving from an electricity grid dominated by fossil fuel generation to a grid with significant solar capacity.

Fig. 1. Solar PV Market Income and Average Wholesale Electricity Prices versus Solar PV Penetration [1]

Figure 1 shows market income for solar plants with increased use of solar. As more solar plants are built, electricity prices at times of high solar output collapse; thus, solar revenue collapses as solar production increases. This limits unsubsidized solar capacity to a relatively small fraction of total electricity production even if there are large decreases in solar capital costs. The same occurs with the large-scale use of wind. This results in increasing number of hours of near-zero wholesale prices for electricity [2, 3] that has been seen in multiple electricity markets in Europe and the United States. This is bad economic news for nuclear, wind and solar for revenue collapse limits the use of low-carbon technologies. It favors lower-capital-cost higher-operating-cost natural gas plants with their capability to cycle up and down quickly. New technologies are required

to store or use this electricity at times of low prices to produce electricity at times of high prices.

We describe two strategies to improve nuclear power economics (Fig. 2) based on heat storage. Heat storage is less expensive than electricity storage [4]. The reactor operates at full capacity all the time but the electricity to the grid varies with time depending upon electricity prices. Each option allows sending heat to industry.

Fig. 2. Alternative Power Reactors with Different Power Cycles and Heat Storage

� Nuclear heat to storage for peak power productionusing light-water reactors (LWRs) and other lowertemperature reactors with steam cycles. The heatcomes from the reactor as steam. The peaktemperature of the stored heat is near the peakreactor coolant temperature. Six classes of heat-storage technologies have been identified—severalthat could be commercialized in less than a decade.Station output can vary from ~30% of base-loadcapacity to several times the base-load capacity ofthe power plant.

� Electricity to high-temperature heat storage forpeak power production using high-temperaturereactors (HTRs) with Brayton power cycles.Electricity is used in a Firebrick Resistance-HeatedEnergy Storage (FIRES) system to heat firebrick totemperatures many hundreds of degrees above thepeak coolant temperature of the reactor. This high-temperature heat is used for an ultra-efficientthermodynamic topping cycle within the Braytonpower cycle. The power output of the station can

863

Transactions of the American Nuclear Society, Vol. 116, San Francisco, California, June 11–15, 2017 Transactions of the American Nuclear Society, Vol. 116, San Francisco, California, June 11–15, 2017

Thermal Energy Storage Systems and their Integration with NPPs

vary from minus several hundred megawatts to plus several hundred megawatts for each 100 megawatts of base-load electricity production; that is, the power plant buys and sells electricity depending upon price while the reactor operates at constant output.

STEAM FROM LWRs TO HEAT STORAGE FOR PEAK ELECTRICITY PRODUCTION

Energy can be stored as heat or work (pumped hydro, batteries, etc.). Nuclear reactors produce heat and heat storage is less expensive than storing work. Integration of thermal energy storage capacity into a nuclear power plant enables variable electricity output from a reactor plant where the reactor operates at base-load. There are several constraints.

� Constant full reactor output. To maximizeeconomics, the reactor should be operated at fullpower. The steam from the reactor can be dividedbetween the main turbine and the storage system.

� Minimum electricity to the grid. For the powerplant to maintain its capability to rapidly send100% of its rated capacity to the grid, the minimumsteam to the turbine is that required for the turbineto remain on-line and be able to rapidly come tofull power by shutting off steam going to storage.That implies the minimum power to the grid is near30%--at which time 70% of the steam is going tothe storage system.

� Maximum electricity to the grid. This is equal to thebase-load capacity of the power plant plus thepower output from the energy storage system. Forsome technologies this output can be 2 to 3 timesthe base-load electricity output.

Six classes of storage options that couple to light-water reactors are being examined [5] where steam is the input to the storage system. For some options, there is the choice to get steam from the storage system that could be fed back to the main reactor turbine if that turbine was oversized.

Steam Accumulators. A steam accumulator is a pressure vessel nearly full of water that is heated to its saturation temperature by steam injection. The heat is stored as high-temperature high-pressure water. When steam is needed, values open and some of the water is flashed to steam that is sent to a turbine producing electricity while the remainder of the water decreases in temperature. Steam accumulators have been used as pressure buffers in steam plants for over a century and have been coupled to solar thermal plants as a mechanism of heat storage. Steam accumulators are capable of rapid charge and discharge cycles. There has been only

limited exploration of the design space for large systems that would couple to a nuclear reactor.

Packed-bed Thermal Energy Storage. A packed-bed thermal energy storage system consists of a pressure vessel filled with solid pebbles with a steam valve at the top and water outlet at the bottom. Heat is stored as sensible heat in the pebbles. To charge the system, steam is injected. The steam condenses as the cold pebbles are heated and water exits from the bottom of the vessel. At the end of the charging cycle all pebbles are hot and there is hot water filling the voids at the bottom of the vessel. To discharge the system, water is injected into the bottom of the vessel and steam is produced by the hot pebbles. Packed beds are more thermodynamically efficient than steam accumulators because they operate in a counter-current mode—the hottest steam sees the hottest pebbles. The window of design options for packed-bed systems, including the range of suitable pebble materials and sizes and the impact of pebble choice on dynamic performance, is only partly explored.

Sensible Heat Fluid Systems. Sensible heat storage involves heating a second fluid with steam, storing that second hot fluid at atmospheric pressure, and using that fluid later to produce steam. This heat storage technology is used with solar thermal systems at temperatures near those of LWRs. A range of fluids have been used in such systems, including oils, and molten nitrate salts. There are two physical configurations: two-tank and thermocline systems. In a two-tank system, one tank will hold cold fluid and one will hold hot fluid, with the ratio of fill levels in the tanks indicating the state of charge. In a thermocline system, hot fluid is injected at the top of the tank, and cold fluid is injected at the bottom. In both cases, one heat exchanger is used to heat the fluid during charging and one is used to cool the fluid to produce steam during discharging. In solar thermal systems two-tank sensible heat storage has been demonstrated at the 100 MWh scale, and the thermocline type has been demonstrated at the 1 MWh scale.

Cryogenic Air Systems. A cryogenic air energy storage system stores energy by liquefying air. A less tightly coupled cryogenic system would use electricity to drive the chilling process; the option exists to more tightly integrate the chilling process with the nuclear plant and use of steam turbines. The liquefied air can be stored in facilities similar to those used to store liquefied natural gas (LNG). The energy storage capacity of the liquid air reservoir can be enhanced through the integration of a sensible heat storage system. To produce electricity, the liquid air is compressed, heated using steam from the NPP secondary side and sent through a gas turbine before being exhausted to the atmosphere—potentially a low-cost peak power cycle. Estimated round-trip efficiency may be as high as 71%. A pilot plant is in operation in the United Kingdom.

Hot Rock Systems. A hot rock energy storage system is similar in concept to a packed bed energy storage system except it operates at atmospheric pressure. A volume of

864

Transactions of the American Nuclear Society, Vol. 116, San Francisco, California, June 11–15, 2017

Thermal Energy Storage Systems and their Integration with NPPs

crushed rock with air ducts at the top and bottom is created. To charge the system, air is heated using a steam-to-air heat exchanger delivering heat from the reactor, then the air is circulated through the crushed rock heating the rock. To discharge the system, the airflow is reversed, and cold air is circulated into the crushed rock. This discharged hot air can be used to (1) produce steam for electricity or industry or (2) hot air for collocated industrial furnaces to reduce natural gas consumption. Large hot rock systems are under development by the shale oil industry (Red Leaf Inc.) to produce oil.

Geologic Heat Storage Systems. Geologic heat storage systems combine the features of an enhanced geothermal energy facility with thermal energy storage. Thermal energy is stored by injecting hot water heated by steam from the surface into the reservoir; energy is discharged by pumping hot fluid back to the surface for electricity production in a conventional geothermal plant. This is the only heat storage option that is a candidate for seasonal energy storage because of the very low cost of the storage media—hot rock.

Each storage technology has different characteristics such as rate of charging, efficiency, cost of storage, and rate of discharge. As a consequence, the preferred option will depend upon the electricity market. The preferred heat storage system in a grid with large solar capacity and the need for daily energy storage will be different than a system with excess wind capacity and multiday cycles of low and high-priced electricity. None of these technologies has yet been deployed with a nuclear system and different systems are at different states of technology development.

ELECTRICITY TO STORED HIGH-TEMPERATURE HEAT FOR A NUCLEAR BRAYTON POWER CYCLE WITH PEAKING CAPABILITY.

This is an advanced option where a high-temperature reactor (HTR) is coupled to an open or closed Brayton power cycle with heat storage. The storage technology is Firebrick Resistance-Heated Energy Storage (FIRES). FIRES can be coupled to industrial furnaces or gas turbines.

MIT is developing FIRES (Fig. 3) to convert excess electricity into stored heat for use in industrial furnaces [6]. When the price of electricity is less than the price of fossil fuels, electricity is used to heat firebrick as high as 1800°C. Air is blown through the firebrick to provide hot air to industrial furnaces as a partial replacement for natural gas or in the future, hydrogen or biofuels. If the hot air is too hot, cold air is added to lower gas temperatures. When electricity prices are low, FIRES is providing hot air to industrial furnaces in addition to the heating of firebrick. FIRES can set a minimum price for electricity near that of natural gas—far above the negative and near-zero wholesale electricity prices found in some electricity markets. The estimated cost of this technology is $5-10/kWh—a factor of 20 to 60 less than projected future costs of batteries. While batteries

provide electricity versus heat, the large differences in capital costs will favor FIRES under most conditions.

Fig. 3. FIRES for Industrial Heat

The longer-term option is incorporating FIRES into a nuclear Brayton cycle coupled to an HTR [7]. Fig. 4 shows one variant—a Nuclear Air-Brayton Combined Cycle (NACC) with FIRES. This variant can be coupled to a Fluoride-salt-cooled High-temperature Reactor (FHR) or a Molten Salt Reactor (MSR). Other variants can couple to HTGRs. NACC power cycles are based on natural gas combined cycles that can operate in two modes: base-load and peak electricity. Nuclear heat is used for base-load electricity production. Additional heat for peak electricity production can be provided by natural gas, stored heat from FIRES or ultimately hydrogen.

Fig. 4. NACC with FIRES

During base-load operation (1) outside air is compressed, (2) heat is added to the compressed air from the reactor through a heat exchanger, (3) the hot compressed air goes through a turbine to produce electricity, (4) the air is reheated and sent through a second turbine to produce added electricity, (5) the warm low-pressure exiting air goes through a heat recovery steam generator (HRSG) to generate steam that is used to produce added electricity and (6) air exits up the stack. The heat-to-electricity efficiency is ~42%.

865

Transactions of the American Nuclear Society, Vol. 116, San Francisco, California, June 11–15, 2017 Transactions of the American Nuclear Society, Vol. 116, San Francisco, California, June 11–15, 2017

Thermal Energy Storage Systems and their Integration with NPPs

The NACC base-load peak temperature is determined by the materials of construction of the reactor coolant-gas turbine heat exchanger. With typical materials, that limit is near 700°C. While these are high temperatures for heat exchangers, they are low temperatures for gas turbines, where there are utility gas turbines with peak temperatures near 1400°C and utility turbines on the test stands with temperatures near 1600°C. Much higher temperatures are possible because gas turbine blades can be cooled from the inside with ceramic coatings on the outside to insulate the turbine blade from the high combustion temperatures.

Consequently, in a gas turbine there is the option of adding heat after the nuclear heating (via heat exchangers) to further raise compressed gas temperatures before entering the second power turbine—a topping cycle. The added high-temperature heat can be provided by natural gas, hydrogen, another combustible fuel or stored heat using FIRES. Our studies used a modified GE 7FB gas turbine—the largest rail transportable gas turbine made by General Electric. Heating the compressed air up to 1065°C results in an incremental heat-to-electricity efficiency of 66.4%. For comparison, the same GE 7FB combined cycle plant running on natural gas has a rated efficiency of 56.9%.

Economic assessments indicate that an FHR with NACC using natural gas will have 50% more revenue in states such as Texas and California than a base-load nuclear plant after paying for the natural gas. The FHR with NACC converts natural gas to electricity with an efficiency of 66.4% versus an efficiency of ~60% for a stand-alone natural gas combined cycle plant and ~40% for a stand-alone natural gas turbine. Using the newest turbines, the incremental natural gas to electricity efficiency is over 70%. The higher efficiency enables NACC to compete with stand-alone natural gas plants.

When FIRES is coupled to NACC, it partly or fully replaces natural gas for peak electricity production. At times of low electricity prices the firebrick is heated using (1) the base-load electricity from the reactor and (2) buying low-price electricity from the grid. Low electricity prices are defined as (1) less than the price of natural gas or (2) more than a 50% difference in electricity prices in a day that makes it worthwhile to convert electricity to heat and back to electricity to increase revenue. The power station both buys and sells electricity depending upon price. The round-trip efficiency is near 66% because the electricity-to-heat efficiency is near 100%. The projected improvements in gas turbines are expected to enable round-trip electricity to heat to electricity efficiencies over 70%--the consequences of a topping cycle combined with advances in gas turbines.

HTRs with FIRES and Brayton power cycles avoid the fundamental weakness of energy storage systems such as hydro pumped storage and batteries. Storage systems do not replace the need for electricity production capacity. Extended hot or cold weather and other conditions can deplete any storage system. At that time one needs electric generating capacity to assure electricity supplies. FIRES can

be depleted but there is the option of using natural gas or other fuels to produce peak power if needed with NACC.

Because of the very high temperatures for efficient gas turbines, only FIRES is capable of providing stored heat. Only firebrick can withstand such high temperatures and only electric heat can generate the required temperatures.

CONCLUSIONS

The transition from a fossil energy world to a low-carbon nuclear-renewable world is a transition from a low-capital-cost high-operating-cost energy systems to a high-capital-cost low-operating-cost energy systems where nuclear, wind and solar must operate at high capacity factors for economic energy production. That implies energy storage. Nuclear power plants convert heat to electricity and thus have the unique capability to couple to heat storage systems where heat storage is less expensive to store than work (electricity). This defines a major direction for nuclear power going forward—base-load reactors with variable energy to the grid and industry.

ACKNOWLEDGEMENT

The authors would like to thank Idaho National Laboratory and the U.S. Department of Energy (DOE) for their support of this work.

REFERENCES

1. MASSACHUSETTS INST. OF TECHNOLOGY. TheFuture of Solar Energy, 2015.

2. MASSACHUESTTS INST. OF TECHNOLOGY,Utility of the Future, 2016.

3. S. BRICK and S. Thernstrom, “Renewables andDecarbonization: Studies of California, Wisconsin, andGermany”, The Electric Journal, 29, 6-12, 2016.

4. X. LUO et al., “Overview of Current Development inElectrical Energy Storage Technologies andApplication Potential in Power System Operation”,Applied Energy, 127, 511-536, 2015

5. D. CURTIS, Shifflet, N., and Forsberg, C. "TechnologyOptions for Thermal Energy Storage with Water CooledReactors", Proceedings of the American NuclearSociety 2017 Annual Meeting, June 11-15, 2017, SanFrancisco, CA, USA.

6. D. STACK, Firebrick Resistance-heated EnergyStorage Systems, MS Thesis, Department of NuclearScience and Engineering, Massachusetts Institute ofTechnology, 2016.

7. C. FORSBERG and P. F. Peterson, “Basis for Fluoride-Salt-Cooled High-Temperature Reactors with Nuclear�Air-Brayton Combined Cycles and Firebrick�Resistance-Heated Energy Storage”, Nuclear�Technology,� ����� ����� �2016��� �������������������������.