Print this article

Comment

1 June 2007

Reaction under pressure

French nuclear power study points to turbine flowmeters for liquid flow sensing to improve safety in pressurized water reactors

By Quentin Grando and Jeff Morrissey

Across the globe, nuclear energy plays a key role in meeting countries' electric power. France is a major producer of nuclear energy, operating nearly 60 nuclear generating stations. It maintains a comprehensive regulatory program governing the operation of nuclear reactors. With help from the Common Access to Biotechnological Resources and Information (CABRI) water loop facility, nuclear power producers can improve the safety and reliability of their pressurized water reactors. The facility's radioactivity-initiated accident (RIA) test device proved to be a challenging instrumentation application, but ultimately, advanced turbine meter technology provided an effective flow measurement solution.

The French government initiated a project to study potential risks of reactivity accidents in pressurized water reactors (PWRs). To verify the safety of nuclear fuel rods, the French regulatory agency, Institut de Radioprotection et de Sûreté Nucléaire (IRSN), performed a series of tests using a pressurized water loop. This environment places extreme demands on all types of instrumentation, including flowmeters subjected to severe pressure spikes, high temperatures, and extreme hydraulic shock loads inside the test device.

IRSN, in collaboration with 14 partners from the U.S., Asia, and Europe, established the CABRI International Program to study the behavior of both uranium dioxide (UO₂) and mixed oxide fuel in the event of an RIA in a PWR. Studies determined the ability of high burn-up fuel to withstand the sharp power peaks occurring in PWRs due to rapid reactivity insertion in the reactor core.

More than 400 nuclear power reactors now operate in 31 countries, producing over 363 billion watts of electricity worldwide, according to the World Nuclear Association. Another 30 reactors are under construction, and 24 countries (including six nuclear reactors not operating) are planning or proposing to build an additional 104 reactors.

Against this backdrop, the industry needs assurance of nuclear energy remaining economically competitive and environmentally friendly. Nuclear energy must also remain acceptable to the public; excellent nuclear safety records are, of course, a prerequisite to this process.

Nuclear reactor process

PWRs are generation II nuclear power reactors using water under high pressure as a coolant and neutron moderator. The primary coolant loop in these reactors is kept under high pressure to prevent the water from boiling. A PWR works because the nuclear fuel in the reactor vessel is engaged in a chain reaction, which heats the water in the primary coolant loop by thermal conduction through the fuel cladding. The hot water is pumped into a steam generator, allowing the primary coolant to heat up the secondary coolant. The steam formed in the steam generator flows through a steam turbine, and the energy extracted by the turbine drives an electric generator. After passing through the turbine, the secondary coolant cools down in a condenser before feeding into the steam generator again. This process reduces pressure at the turbine outlet, thus improving thermal efficiency.

The uranium used in PWR fuel is usually enriched several percent in ²³⁵U. After enrichment, the UO₂ powder fires in a high-temperature sintering furnace to create hard, ceramic pellets of enriched UO₂. The pellets then go into tubes of a corrosion-resistant zirconium metal alloy (Zircoloy), which backfill with helium to aid heat conduction. The finished fuel rods group in fuel assemblies, called fuel bundles, used to build the core of the reactor. A typical PWR has fuel assemblies of 200 to 300 rods each, and a large reactor can have 150-250 such assemblies with 80-100 tons of uranium in all.

In a PWR, it is normal to fill the gap between the fuel rods and cladding with helium gas to optimize thermal contact. During reactor operation, the amount of gas inside the fuel pin can increase because of the formation of noble gases (krypton and xenon) by the fission process. If an RIA occurs, the temperature of this gas can increase. Since the fuel pin is sealed, the pressure of the gas will increase (P = nRT/V), and it is possible to deform and burst the cladding.

Fuel testing procedure

As part of the CABRI project, IRSN measures phenomena occurring before, during, and after an RIA. Engineers at an experimental reactor operated by the French Atomic Energy Commission at Cadarache, in southern France, use an RIA test device with water and sodium loops to perform a series of tests on irradiated combustible nuclear rods in a representative thermo-hydraulic environment.

The RIA test device, the only one of its kind in the world, enables researchers to accurately determine the thermal-hydraulic and neutronic conditions of sample fuel rods. The design of this device prevents any damage to the in-pile cell resulting from pressure loading due to possible fuel-coolant interaction (FCI). It also ensures the containment of fuel particles during experiments.

Nuclear facilities submit their fuel rods to IRSN for evaluation under a simulated worst-case scenario-they remove the moderator material (neutron absorbing) from the test device. They conduct tests under conditions typical of a PWR core, raising and lowering moderator rods into the reactor to control reaction rate. Under certain circumstances, these rods can bend and are unable to slide into the reactor. Also, the mechanical device used to raise and lower the rods is subject to failure.

Testers place the fuel rod sample into the test channel and use a pulse input to start a nuclear chain reaction (pulse input = accelerated neutron). The power transient has a pulse width (full width at half maximum) of 28 milliseconds and deposits energy of 89 cal/g in the test fuel. Removing the moderating material (Helium 3) begins the process of simulating an RIA. If the rod cracks during testing, it signifies failure, and the plant must replace the existing rods. If the rod passes, then the facility may continue to use its current material for a specified period of time.

Instrumentation requirements

The RIA test device holds a variety of instruments to measure transient conditions such as temperature, pressure, sound, and flow. This includes ultrasonic sensors for cell water level measurement as well as pressure sensors for fuel rod and channel pressure measurement. Due to the severe nature of the water-loop environment, however, IRSN found it difficult to locate suppliers who could provide equipment to meet its demanding requirements.

The RIA test channel characteristics include high temperature, pressure, and neutronic flux. The physics of the test operation dictate measurement instruments with a very fast response time. In addition, space is limited for mounting sensors in the test channel; channel walls are thick to resist against high-pressure loading FCI causes.

Flow measurement was a particularly challenging problem for the CABRI facility. IRSN needed an accurate flow-metering device for measuring de-mineralized water under initial low-flow, semi-state conditions-helping technicians calculate heat transfer between the reactor and test channel. Flow rates can vary from 0.1 to 2 m^3/hr under normal testing conditions. In the event of a fuel rod failure, flowmeters must withstand hydraulic shock caused by flow rate acceleration of 0.1 to 10 m^3/hr over a duration of 1 milliseconds. Depending on meter location in the test channel, shock waves can be in the forward or reverse flow direction.

IRSN required a strong, rugged flowmeter able to handle the rigors of its simulated reactor environment. In case of an explosion or other accident, no part of the meter could disperse into the test channel. The device would also have to work under specific radiation conditions in the flow measurement area, including neutronic flow (instantaneous: 10^E 13 n/cm²/s, integrated: 2.10^E 17 n/cm²) and gamma flow (instantaneous: 50 Gray/s, integrated: 10^E5 Gray).

Most important, the design of the CABRI facility required welding the flowmeter in place so it would fit within a 60 mm diameter test channel area. This size requirement eliminated the use of most flow measuring technologies.

Flow measurement solution

IRSN engineers found one type of instrument, a turbine flowmeter, met all its project requirements. The meter's design incorporates a freely suspended rotor turned by fluid flow through the meter body. Since the flow passage is fixed, the rotor's rotational speed is a true representation of the volumetric flow rate. The rotation produces a train of electrical pulses, which an external pickoff senses, counts, and totals. The number of pulses counted for a given period of time is directly proportional to flow volume.

Applications for turbine meters involve measuring low viscosity fluids in a process that requires local and remote indication of rate and total. Turbines are capable of accurately measuring liquids with higher viscosities, as long they remain relatively stable.

High-shock meter design

Due to high pressure in the RIA test channel, IRSN engineers specified a heavy-duty turbine flowmeter configuration. High-shock meters specifically designed for the CABRI project can handle severe pressure spikes, as well as the "water hammer" effect, which may cause damage to conventional flow sensors. Water hammer (or, more generally, fluid hammer) is a pressure surge or wave caused by the kinetic energy of a fluid in motion when it is forced to suddenly accelerate, stop, or change direction.

Based on the high-shock specifications, turbine flowmeters have 316L stainless steel housings and corrosion-resistant tungsten carbide journal bearings to withstand high temperature and shock conditions. The journal bearings can handle shock loads in both forward and reverse flow directions. The meter's internal components are captivated by tube extensions with a smaller bore to prevent hydraulic shock pushing them through.

The turbine meter housing undergoes machining from a single block of 316 SS, with an inner bore for the flow and two cavities: one for inserting the pickoff sensor; the other for holding a small junction box that allows joining securely the pickoff wires and high-temperature cable assembly. The cable assembly consists of nickel conductors, alumina insulator, and stainless-steel sheath. After meter assembly, the cavities fill with a high-temperature ceramic potting, and a cover plate seals them.

The high-shock meter rotor blades are thicker than those on a standard turbine meter. Machining a radius at the intersection of the blades and rotor hub eliminates the stress point created by a 90-degree cut. A step is also located in the upstream bore of the rotor so it cannot be pushed over the bearing.

When space is limited, the small size of the flowmeter allowed the supplier to develop meters with a special low-profile housing that includes an integral pickoff sensor coil and a separate chamber to hold the cable assembly termination. They welded a section of straight pipe onto the upstream and downstream ends of the housing. Constant temperature monitoring, heat sinks, and argon purging prevent damage or contamination to the meter internals during the welding process. The straight pipe extensions have a dual purpose; they allow IRSN to weld the flowmeters into their test channel, and they hold the meter's internal components securely in place. The extensions have machined grooves in the ends to match the outer end of the internal supports and keep the internals from rotating.

Turbine flowmeters installed on the RIA test device have a standard flow range of 0.44 to 8.8 GPM, and accuracy of better than 1% of reading. Although flow measurements are not required during control rod ejection, the meters can operate in temperatures up to 350ºC (661ºF) and pressure up to 400 bar (5800 PSIG), as well as flow rates from 1 to 10 m3/h (4.4 to 44 GPM).

To meet RIA testing and documentation requirements, each instrument is subjected to a hydrostatic pressure test, helium leak test, radiographic test, dye-penetrant test, insulation resistance test, and high-temperature function test. The meters mount vertically with flow going up during calibration to simulate their actual orientation during nuclear fuel rod test sequences. It is also helpful to provide the user with test procedures, reports, and acceptance criteria for testing; material certificates for all meter components; and welding specifications and welder qualifications.

About the Authors

Quentin Grando (quentin.grando@irsn.fr) is a test engineer at Institut de Radioprotection et de Sûreté Nucléaire (IRSN) in Cadarache, France. Jeff Morrissey (Morrissey@ftimeters.com) is an applications engineer at Flow Technology, Inc., in Phoenix.