BACKUP ELECTRICITY SUPPLY SYSTEMS ARE required to provide cooling for decay heat produced in shut-down reactors and to maintain services (eg system control, lighting, communication, and ventilation) to the reactor, turbine generator and ancillary plant. The electricity required by backup systems will vary with reactor size, efficiency, design, spent fuel and waste management system.

The question for safety authorities and operators is, ‘how long will backup energy supply be required in the worst event scenario?’ The answer will require risk/ hazard analyses, coupled with knowledge gained since Fukushima and a little conservative stargazing. A second concern, if battery storage is used, is that the electricity going into storage units will be alternating current (AC) while that exiting storage will be direct current (DC), so the performance and reliability of rectifiers, inverters and very high-speed switches must also be assured.

Backup and auxiliary power supply can take various forms. A connection to the power grid is the obvious immediate choice if the connection is operational and can call on reliable dispatchable power. The second backup is usually generators fired with diesel, light fuel oil or other suitable fuels. There are redundancies including multiple sets and twinned sets, sometimes separated by distance. Gas turbine sets can replace reciprocating diesel engine sets in some instances.

There are batteries to cold-start the diesel generators and power control, security and communication systems (alternatively, energy for cold starts can be supplied by compressed air).

For over a century the primary rechargeable industrial batteries were lead-acid type. They are still the principal energy storage for operating submerged diesel-electric submarines and were the only realistic battery technology that could be applied to powering electric vehicles and starting emergency diesel generators. There may be better options on the way. Battery systems being developed have greater power density (energy stored per unit of mass or volume), faster recharge, greater charge and discharge efficiencies (wasting less heat energy), lower self-discharge rates, improved shelf-lives, an improved constant peak voltage discharge to time relationship, improved cycle durability, very low thermal runaway propensity and low fire and explosion hazard potential. They are likely to replace the lead-acid batteries used in limited capacities in nuclear plants and may replace or supplement diesel generators to supply emergency power for short periods, such as a few hours after an emergency shut down.

Power supply units that have rotating mechanical parts, take seconds to start up and supply power, and require external fuel (of guaranteed quality). Combustion air and active cooling systems are not appropriate as the primary power backup systems for nuclear plants. The Fukushima accident showed that diesel-fired standby generators could be overwhelmed by flooding.

New battery technologies may have advantages over diesel generators that should be investigated and quantified. If they prove to be superior, they should be used in existing and future nuclear plants.

Battery technology has greatly advanced over the last twenty years. The drivers have been devices like laptop computers and mobile phones, and more recently electric vehicles that can compete in range and energy efficiency with fuelled vehicles. New battery technologies are still developing and it is too early to say which will be the winners and losers for specific applications, including heavy duty power standby units for critical infrastructure such as nuclear plants.

Batteries for nuclear plants should offer: secure dispatchable power output; very high charge and discharge cycle performance; security from uncontrolled cascading discharge; zero gas discharge; low internal charge and discharge resistance; passive cooling; no moving parts (eg electrolyte circulation); acceptable energy efficiency, specific energy and energy density; and acceptable cost.

On a multi-billion dollar nuclear plant the cost of a primary hazard reduction system should not be a governing factor in its construction, operation or maintenance.

The battery backup system must be robust. The Fukushima earthquake and tsunamis made it obvious that supporting systems, access, transmission line integrity and communications were not robust. Relying on business-as- usual access and communications is potentially a major hazard.

Batteries, diesel generators and outside sources should be noted and included in emergency plans however they should not be the core (including grid back-up or bypass lines and private gen sets, as well as dedicated diesel gensets and regional backup batteries). Regular safety exercises to identify new technologies can offer new options. Regular hazard assessments should include analysis of the power backup systems, including how each level of backup comes in (and goes out).

Cooling and other power use systems for nuclear power

Current backup systems (limited battery supply, linking back to the grid and on-site diesel generation) provide a full dispatchable and variable supply to meet the demand of the reactor and ancillary services.

In some cases diesel reciprocating engines can be replaced by gas turbine sets, but the latter have longer start-up and full-power times. In the case of rapid start-up they require volatile fuels.

Some of the fuels used for diesel engine sets include natural gas supplemented with diesel and condensate; and some such fuels should not be used on a nuclear site due to their high flammability. Gas turbines should only have a role in emergency generation if they are located at a safe distance outside the perimeter.

New nuclear power stations are likely to be close to seawater cooling. Alternatives are air cooling, freshwater cooling and hybrid cooling (air cooling boosted by an auxiliary water-cooling), but in tropical, sub-tropical and many temperate situations, to achieve acceptable cooling efficiencies high water flux direct cooling provided by seawater is required. Post Fukushima new fail-safe cooling systems are required.

Nuclear units with isotope production and research facilities also require uninterruptable power; these units plus their administration divisions should be linked with the power plant’s backup power supply.

Desirable battery characteristics

A battery used for nuclear power plant backup must be able to supply its designed emergency power (MW) and energy (MWh) quickly (less than 10s to full power), without significant deviation in performance over long periods of time and in the event of multiple demand events. The batteries must be fully rechargeable no matter what their initial charge level is before recharging — even if at zero charge prior to charging. New batteries will likely have very low or zero ‘dead charge’ (ie charge that cannot be accessed).

Among nuclear facilities’ safety regulations are those referring to the threat of fire or heating. Some battery types have a propensity to chemically cascade and get hotter than would be within normal operational heating ranges. Those batteries can catch fire and any fire in a nuclear facility is a reportable incident. Lithium-ion batteries are
an example. It is suggested that Li-ion batteries should not be used in IAEA regulated facilities; this should encompass computer batteries, safety illumination and uninterruptable power supplies (UPS) used in small-scale DC control systems.

Battery basics, configuration, management and placement

Battery cells have an indicative voltage and maximum stored energy along with charging and discharging patterns. During charging and discharging they require heat rejection and they are arranged in banks that allow for passive cooling. In some instances, the banks will be hydraulically sealed and capable of immersion where flooding is possible.

Cells sized at (for example) 3V will be electrically linked in battery modules that have voltages or energy delivery appropriate to requirements. The cells should be fully modular, capable of individual analysis and selectively replaceable, and be arranged in battery packs that allow passive cooling. When interconnected the battery packs may be considered as a single battery but located both inside and outside the plant perimeter. Cell health monitoring systems will be required to provide assurance of capacity and performance during use.

The power generated and dispatched by nuclear plants is high voltage AC; batteries are DC. Where batteries are used as a backup power conversion technologies are required.

Modern technology for converting between DC and AC uses large-scale solid-state electronics that is very reliable. But the operability of the DC system — batteries, inverters, reticulation cabling, switching and monitoring systems — will have to be tested and recorded at regular intervals. Diesels are being produced with a guaranteed start to full load capability of ≤10s; batteries with suitable switching can do the same in nanoseconds, so for an ‘uninterruptable’ power supply, batteries are superior to generators.

But both have challenges. In the case of gensets they are:

Fuel quality (including contamination) and age, which is a significant challenge for diesel plant. ‘Old fuel’ should be recycled back to suppliers or used in testing.

Start-up time. As above, gas turbines can start up more quickly than diesel gensets, but their volatile fuel should not be stored or used close to the nuclear unit. It is worth remembering that nuclear plants have both a physical and an ‘inferred’ perimeter. The physical delineates the area that includes the reactor, generation plant, fuel-storage, services, site management and logistics and security. This area comes under nuclear safety treaties, legislation and regulation. But if backup emergency power sources are outside the physical perimeter they and their connectivity are still an integral part of it; they are ‘inferred plant’ with respect to those treaties, legislation and regulation.

Fuel supply. If emergency diesel generators are 38% efficient and fired with diesel to serve one nuclear unit, the diesel demand will be around 0·425litre/s, so over 48 hours diesel demand would be 172,000 litres. Diesel sets comprise chemical, mechanical and electrical systems. Moving parts wear, fail and have control system problems. US nuclear standards suggest that failure rate must be less than 1% of start-ups.

Robustness. At Fukushima, the diesel generators had first to handle the vertical and horizontal motion and shockwaves and were later inundated with seawater.

In the case of batteries, the challenges are:

Connection. Batteries need a grid link to recharge, possibly 6·6kV networks rather that the existing 230/240kV connection.

Heat. As above, large stationary sealed battery packs are required that can expel heat in a passive manner and can survive shaking and inundation. Battery passive cooling, even at the expense of cooling efficiency with respect to time required for recharge, may become a major factor in battery selection. In some cases (see Figure 2) there is space between battery modules that could contain air, lightly compressed nitrogen or helium or conductive liquids for heat expulsion. Heat removal would be accomplished by convection inside the battery pack shell and heat convection and low temperature radiance externally.

Water-safe. If the batteries are liable for immersion then sealed batteries would be required with adequate heat rejection both when dry and when immersed.

Charging and recharging. Solid-state components such as rectifiers and inverters must also be able to survive an event. Frequency and depth of testing and maintenance, discharge exercises to ensure that each battery pack (battery component) is fully operational and ready for meeting emergency needs, and protection of valuable safety and security assets are required of nuclear power plant batteries. Note that electric vehicle batteries may not be suitable as they require the maximum energy density and specific power possible to give vehicles acceptable range and acceleration. Energy densities and specific power are not as crucial for large stationary batteries backing up nuclear plant.

Complementary technologies

But batteries and generators are not mutually exclusive. Battery requirement to cover the initial four hours of the power failure at a 1300MWe plant would be 20MWh (15% of the Tesla battery array in South Australia). But following initial supply from the battery the diesel could cut in, take up the load from the reactor and begin battery recharge.

Developments in large stationary industrial batteries may provide the nuclear power industry with a new source of backup power in a more thorough and secure way than was demonstrated by diesel generators during the Fukushima disaster. Battery recharge from diesel should be a feature of new and refurbished backup systems; the recharge (followed by discharge) would be a monitoring and maintenance procedure. Battery recharge from diesels with surplus capacity during emergencies should be a built-in option than will assist in following the demand and supply pattern and providing power for unexpected secondary disruptions.

Depending on how total backup is configured, having part of the backup outside the perimeter may reduce the hazard level. Having sufficient capacity inside the perimeter to manage decay heat is important for hazard reduction measures; if that capacity was in the form of battery backup that could be available in nanoseconds and last for several hours, all the better.

In conclusion: we should aim to move diesel sets from Backup Level 1 to Backup Level 2, if new battery technologies can provide additional power security.


Dr Mike Clarke has qualifications in mining and chemical engineering and risk management (CPEng, FIEAust, FAusIMM) and expertise and experience in hazard identification and risk reduction and environmental engineering. Consulting to the energy industries is a major focus