Information Papers

Nuclear Desalination

(November 2008)

It is estimated that one fifth of the world's population does not have access to safe drinking water, and that this proportion will increase due to population growth relative to water resources. The worst-affected areas are the arid and semiarid regions of Asia and North Africa. Wars over access to water, not simply energy and mineral resources, are conceivable.

Fresh water is a major priority in sustainable development. Where it cannot be obtained from streams and aquifers, desalination of seawater or mineralised groundwater is required.  An IAEA study in 2006 showed that 2.3 billion people live in water-stressed areas, 1.7 billion of them having access to less than 1000 m3 of potable water per year.  With population growth, these figures will increase substantially.  Further demand in the longer term will come from the need to make hydrogen from water.

Desalination

Most desalination today uses fossil fuels, and thus contributes to increased levels of greenhouse gases. Total world capacity is approaching 40 million m³/day of potable water, in some 15,000 plants. Most of these are in the Middle East and north Africa, using distillation processes.  The largest plant produces 454,000 m³/day.

The major technology in use and being built today is reverse osmosis (RO) driven by electric pumps which force water through a membrane against its osmotic pressure*.  Multi-stage flash (MSF) distillation process using steam was earlier prominent and it is capable of using waste heat from power plants.  With brackish water, RO is much more cost-effective, though MSF gives purer water than RO.  A minority of plants use multi-effect distillation (MED) or vapour compression (VC).  MSF-RO hybrid plants exploit the best features of each technology for different quality products.

*  About 27 Bar.  Therefore RO needs compression of much more than this.

Desalination is energy-intensive. Reverse Osmosis needs about 4-6 kWh of electricity per cubic metre of water (depending on its salt content), hence 4MWe will produce about 1000 m3 per day.   MSF and MED require heat at 70-130°C and use 25-200 kWh/m³,  though a newer version of MED (MED-MVC) is reported at 10 kWh/m3 and competitive with RO.  A variety of low-temperature heat sources may be used, including solar energy. For brackish water and reclamation of municipal wastewater RO requires only about 1 kWh/m3.  The choice of process generally depends on the relative economic values of fresh water and particular fuels, and whether cogeneration is a possibility.

Some 10% of Israel's water is desalinated, and one large RO plant provides water at 50 cents per cubic metre. Malta gets two thirds of its potable water from RO. Singapore in 2005 commissioned a large RO plant supplying 136,000 m³/day - 10% of needs, at 49 cents US per cubic metre.

Small and medium sized nuclear reactors are suitable for desalination, often with cogeneration of electricity using low-pressure steam from the turbine and hot sea water feed from the final cooling system. The main opportunities for nuclear plants have been identified as the 80-100,000 m³/day and 200-500,000 m³/day ranges.

A 2006 IAEA report based on country case studies showed that costs would be in the range $0.5 to $0.94 /m3 for RO, $0.6 to $0.96/m3 for MED and $1.18 to 1.48/m3 for MSF processes, with marked economies of scale.  Nuclear power was very competitive at today's gas and oil prices.  A French study for Tunisia compared four nuclear power options with combined cycle gas turbine and found that nuclear desalination costs were about half those of the gas plant for MED technology and about one third less for RO.  With all energy sources, desalination costs with RO were lower than MED costs.

Desalination: nuclear experience

The feasibility of integrated nuclear desalination plants has been proven with over 150 reactor-years of experience, chiefly in Kazakhstan, India and Japan.

The BN-350 fast reactor at Aktau, in Kazakhstan, successfully produced up to 135 MWe of electricity and 80,000 m³/day of potable water over some 27 years, about 60% of its power being used for heat and desalination. The plant was designed as 1000 MWt but never operated at more than 750 MWt, but it established the feasibility and reliability of such cogeneration plants. (In fact, oil/gas boilers were used in conjunction with it, and total desalination capacity through ten MED units was 120,000 m³/day.)

In Japan, some ten desalination facilities linked to pressurised water reactors operating for electricity production have yielded 1000-3000 m³/day each of potable water, and over 100 reactor-years of experience have accrued. MSF was initially employed, but MED and RO have been found more efficient there. The water is used for the reactors' own cooling systems.

India has been engaged in desalination research since the 1970s and in 2002 set up a demonstration plant coupled to twin 170 MWe nuclear power reactors (PHWR) at the Madras Atomic Power Station, Kalpakkam, in southeast India. This Nuclear Desalination Demonstration Project is a hybrid reverse osmosis / multi-stage flash plant, the RO with 1800 m³/day capacity and the MSF 4500 m³/day costing about 25% more. They incur a 4 MWe loss in power from the plant.

Much relevant experience comes from nuclear plants in Russia, Eastern Europe and Canada where district heating is a by-product.

Large-scale deployment of nuclear desalination on a commercial basis will depend primarily on economic factors. The UN's International Atomic Energy Agency (IAEA) is fostering research and collaboration on the issue, and more than 20 countries are involved.

One obvious strategy is to use power reactors which run at full capacity, but with all the electricity applied to meeting grid load when that is high and part of it to drive pumps for RO desalination when the grid demand is low.

New projects

South Korea has developed a small nuclear reactor design for cogeneration of 90 MWe of electricity and potable water at 40,000 m³/day. The 330 MWt SMART (System integrated Modular Advanced Reactor) reactor (an integral PWR) has a long design life and needs refuelling only every 3 years. The feasibility of building a cogeneration unit employing MSF desalination technology for Madura Island in Indonesia is being studied. Another concept has the SMART reactor coupled to four MED units, each with thermal-vapour compressor (MED-TVC) and producing total 40,000 m³/day.

Spain is building 20 RO plants in the southeast to supply over 1% of the country's water.

In the UK, a 150,000 m³/day RO plant is proposed for the lower Thames estuary, utilising brackish water.

In India plants delivering 45,000 m³ per day are envisaged, using both MSF and RO desalination technology.

China is looking at the feasibility of a nuclear seawater desalination plant in the Yantai area producing 160,000 m³/day by MED process, using a 200 MWt reactor.

Russia has embarked on a nuclear desalination project using dual barge-mounted KLT-40 marine reactors (each 150 MWt) and Canadian RO technology to produce potable water.

Pakistan is developing a demonstration MED desalination plant coupled to its KANUPP reactor (125 MWe PHWR) near Karachi to produce 1600 m3/day (it was earlier projected to produce three times this).  It has been operating a 454 m3/day RO plant for its own use.

Tunisia is looking at the feasibility of a cogeneration (electricity-desalination) plant in the southeast of the country, treating slightly saline groundwater.

Morocco has completed a pre-project study with China, at Tan-Tan on the Atlantic coast, using a 10 MWt heating reactor which produces 8000 m³/day of potable water by distillation (MED).  The government has plans for building an initial nuclear power plant in 2016-17 at Sidi Boulbra, and Atomstroyexport is assisting with feasibility studies for this.

Egypt has undertaken a feasibility study for a cogeneration plant for electricity and potable water at El-Dabaa, on the Mediterranean coast. Late in 2008 plans were being formed for two 1000 MWe reactors to be built there by 2017-18.

Algeria is considering a 150,000 m³/day MSF desalination plant for its second-largest town, Oran (though nuclear power is not a prime contender for this).

Libya: in mid 2007 a memorandum of understanding was signed with France related to building a mid-sized nuclear plant for seawater desalination.  Areva TA would supply this.

Iran: A 200,000 m³/day MSF desalination plant was designed for operation with the Bushehr nuclear power plant in Iran in 1977, but appears to have lapsed due to prolonged construction delays.

Qatar has been considering nuclear power and desalination for its needs which are expected to reach 1.3 million m3 per day in 2010.

Jordan has a "water deficit" of about 1.4 million m3 per day and is actively looking at nuclear power to address this, as well as supplying electricity.

Argentina has also developed a small nuclear reactor design for cogeneration or desalination alone - the 100 MWt CAREM (an integral PWR).

Most or all these have requested technical assistance from IAEA under its technical cooperation project on nuclear power and desalination. A coordinated research project initiated in 1998 reviewed reactor designs intended for coupling with desalination systems as well as advanced desalination technologies. Safety and reliability are key requirements. This program is expected to enable further cost reductions of nuclear desalination.

Sources:
IAEA 1997, Nuclear Desalination of Sea Water, proceedings of 1997 Symposium.
IAEA 1998, Nuclear heat applications: design aspects and operating experience, IAEA-TECDOC-1056.
Konishi & Misra, Freshwater from the Seas, IAEA Bulletin 43, 2; 2001.
IAEA Nuclear Desalination, paper on web.
International J of Nuclear Desalination, 2003, vol 1, 1.
UN World Water Development Report 2003.
New Scientist 10/7/04.

Seneviratne, G 2007, Research projects show nuclear desalination economical, Nuclear News April 2007.