PWR reactors must be designed to handle any primary loop failure, whether small or large, because the reactor core will melt if not continuously cooled. High-pressure safety injection (HPSI) pumps keep the core flooded and cooled in the initial stages of an accident. This article explains how such pumps are tested to ensure they can be relied on to meet this critical safety function.

Frank Costanzo and Mike Eftychiou, Flowserve, USA

In a pressurized water reactor (PWR), heat generated by the nuclear reaction is transferred to steam generators by reactor coolant pumps via the primary loop piping.

A failure of the primary loop piping is referred to as a loss-of-coolant accident (LOCA). PWR reactors must be designed to handle any primary loop failure, whether small or large, because the core of the reactor will melt if it is not continuously cooled.

High-pressure safety injection (HPSI) pumps are used to keep the core flooded and cooled in the initial stages of a LOCA. Accordingly, these pumps are a key part of one of the most important safety systems in a PWR power plant.

High-pressure safety injection pumps

The HPSI pump is usually a multistage centrifugal type that has a performance curve with a constantly rising head to shutoff (zero flow). The pumps feature face-type mechanical seals at both the inboard and outboard ends, and use anti-friction radial and thrust bearings within a self-contained bearing housing.

Nuclear regulatory requirements in all countries specify that any equipment important to the safe shutdown of a nuclear reactor be qualified to meet its specified functional requirements. This is typically done by engineering analysis and/or testing. Nuclear power plant operators can sometimes insist that analytical methods be supported by actual test results, and original equipment manufacturers periodically choose to verify their analytical models through testing.


Figure 1. Thermal data from a single-loop temperature transient test.
Click here to enlarge image

During a LOCA, the HPSI pumps are subject to severe fluid temperature transients, and the pumps must be properly designed and qualified for this eventuality. Specific design characteristics to be considered include rotor dynamic stability, internal running clearances, coefficient of thermal expansion between mating components, surface hardness and thermal distortion of the pump housing. Failure to consider design aspects could result in metal-to-metal contact, material galling, pump rotor instability or insufficient running clearances.

A nuclear reactor operates at roughly 260 ºC, while the water storage tanks that initially supply the HPSI pumps are kept at an ambient temperature. Once the storage tanks are exhausted, HPSI pumps start taking suction out of the sump below the reactor vessel where the hot leaking reactor coolant is collected, shocking the HPSI pump with a rapid increase in temperature. This extreme temperature change, which can occur in as little as five seconds, puts stress on the HPSI pump.

The impact of these temperature transients can be determined using modern analytical tools, but it is sometimes necessary or desirable to perform a temperature transient testing (TTT) to simulate the event on a full-scale pump.

An innovative way to perform TTT on HPSI pumps is by using a single-loop configuration rather than the typical double-loop configuration. This innovative single-loop test design is simpler and less costly to perform, reducing overall project costs.

Single-Loop Temperature Transient Testing

The innovative single-loop TTT involves a closed loop containing a storage tank to provide a source of water at the specified final temperature for the test (see Figure 2, above). A double-loop configuration includes two setups, with the valve switching between the hot and cold loop. This is a less efficient method compared with a single-loop setup, where a cold tank is connected to a hot loop, and vice versa.


Figure 2. Single-loop temperature transient testing set up.
Click here to enlarge image

For each of the two possible thermal transients – cold-to-hot and hot-to-cold –- the HPSI pump must be maintained at the initial transient temperature while the final transient temperature is achieved in a storage tank. During the cold-to-hot transient, temperatures typically increase from 4 ºC to 65 ºC. A typical hot-to-cold transient involves a temperature decrease from 49 ºC to 4 ºC. These temperature changes must happen in five seconds to simulate what an HPSI pump might encounter during the safe shutdown of a reactor.

After the temperature stabilizes, an isolation tank between the pump and supply tank is opened. Simultaneously, a drive motor is energized to bring the pump to the desired flow rate.

To accomplish the cold-to-hot thermal transient, which is the most severe, the HPSI pump and any water within it is cooled to 4 ºC through a temperature chamber surrounding the pump. Simultaneously, the test-stand water supply is heated to the prescribed temperature of 65 ºC.

After temperatures have stabilized, the cold pump is started and the hot water enters the pump within the time defined by equipment specifications, which is typically five seconds. The pump must then be operated for one hour at the elevated fluid temperature.

The hot-to-cold thermal transient is conducted in a similar, but reverse, manner. The HPSI pump chamber is heated to a maximum temperature of 49 ºC, while the test-stand water supply is cooled to 4 ºC. After temperatures have stabilized, the pump is energized and the transient is implemented within five seconds.

This test is much more onerous for pumps with many internal parts. HPSI pumps must be able to meet reactor pressure; therefore, HPSI pumps are 12-stage, double-case pumps, which have more parts than low-pressure single-stage pumps and must have an especially robust design to withstand the test.

If the HPSI pump design is able to successfully withstand the harshest cold-to-hot and hot-to-cold thermal transients it might encounter, plant operators can be confident of reliable performance during an LOCA.

For a TTT to be considered successful, several conditions must be met:

  • The HPSI pump must operate without incident throughout the duration of both tests and for at least one hour after the tests are complete.
  • Vibration levels must stay within specified limits to ensure no damage occurs to the pump’s internal components.
  • The electrical power demand cannot exceed the motor service factor from the rated power for more than one minute. Increases in motor power are indicative of internal pump rubs or contact between rotating and stationary components, which cause thermal distortion or binding of the volute casing and pump rotor.

A post-test disassembly of the HPSI pump determines if any internal parts have dislodged or made contact during the course of the TTT. First, the volute assembly is removed from the barrel and the volute halves separated to permit inspection of the rotating element. Next, inspection of the inside portions of the outer barrel, the volute casing and the rotating element must show no evidence that any part or component is has come loose or is damaged.

Results of TTT

The simplified single-loop TTT methodology reduces project costs for nuclear plants while ensuring HPSI pumps can be relied on to meet their critical safety function.

A centrifugal pump design such as the HPSI is complex. As a result, it should be configured to conform to a comprehensive set of design and operating requirements through a systematic approach and methodology to evaluate plant system operation and transients in conjunction with ASME code requirements.

About the Authors:

Frank Costanzo was the director of engineering at Flowserve Corp’s Vernon operations (recently retired). He holds a BS in mechanical engineering and is a US Department of Energy certified pump specialist.

Michael Eftychiou is engineering section head at Flowserve Corp’s Vernon nuclear products operations. He holds a BS in mechanical engineering.