George Antaki, PE
Nuclear power plants are designed, constructed, operated and inspected to prevent pipe ruptures. Despite these preventive measures, plants are designed to mitigate the effects of hypothetical pipe ruptures. The hypothetical ruptures are in the form of leakage cracks and, for high energy lines (operating temperatures in excess of 200 degrees F or operating pressures in excess of 275 psig), full circumferential and longitudinal breaks. Nuclear power plants are designed to achieve a safe shutdown if such a postulated high energy line break (HELB) was to occur.
The methods and criteria for HELB analysis are defined in Section 3.6 of the NRC Standard Review Plan (NUREG 0800). There are basically two consequences to the postulated HELB:
(1) Dynamic effects in the form of pipe whip, jet impingement on surrounding targets, in-pipe fluid transients (waterhammer) caused by the sudden break, and sub-compartment pressurization due to the discharge of hot pressurized fluid inside rooms and compartments, and
(2) Environmental effects in the form of flooding and spray wetting from fluid discharge from the break, and harsh ambient temperatures and humidity. Challenged by these effects, plant systems and components must be designed to operate to bring the plant to a safe shutdown.
To limit and confine the dynamic effects of postulated HELB, plants install whip restraints, bumpers, jet shields and barriers. These protective structures are designed using energy methods or dynamic analysis techniques, including finite element analysis, such as illustrated in the figure.
Figure – Analysis Model of Pipe Whip Against a Crushable Bumper
The interest in pipe breaks originated with the need to size the emergency core cooling systems (ECCS), when it became important to understand break size, shape, opening area and opening time. The analysis of high energy line breaks then evolved to encompass dynamic effects such as pipe whip and jet impingement, and the regulatory expectations were spelled-out in 1974, in Regulatory Guide 1.46, later to be replaced by Standard Review Plan Section 3.6.
The nuclear power industry has studied the dynamic effects of postulated HELB for the past 45 years. These studies have included theoretical, experimental, and numerical research. The pioneering works in the field included a 1965 study on two-phase blowdown from pipes and a 1969 report on blowdown and jet forces, both by Dr. Moody, and the early experimental work by Faletti and Moulton in 1969. This was to be followed by more experiments aimed at improving our understanding of the transient phenomena that take place when a break occurs in a high pressure and high temperature pipe. Following is a quick overview of some of the experimental work conducted since the early 1970’s.
In 1975 Tractionel (Belgium) conducted tests on the energy absorbing capabilities of stainless steel U-bars, compression copper bumpers, and cellular concrete.
From 1974 through 1986 the Japan Atomic Energy Research Institute (JAERI) conducted a series of jet discharge tests and studies on 4 inch, 6 inch, and 8 inch pipe under PWR and BWR primary loop pressures and temperatures. The tests were also instrumented to measure target temperature.
In 1979, Westinghouse conducts tests on pipes used as elements of pipe whip restraints.
In 1980 and 1981, Studsvik conducts large scale jet impingement tests at the Marviken power plant in Sweden. The pipe sizes tested ranged from 8 inch to 20 inch, with pressures up to 700 psi, and fluid ranging from subcooled water to steam.
In the mid-1980’s JAERI also conducted tests of pipe whip with U-bar whip restraints. The whipping pipes were 4 inch and 6 inch, pressurized at BWR as well as PWR pressures and temperatures. The tests investigated the dynamics of the whipping pipe and the formation of a hinge at the whip restraint.
Figure - Recent FEA Benchmarked Against the JAERI tests
In the mid-1980’s the Atomic Energy of Canada and the Electric Power Development Company of Japan conducted pipe-on-pipe tests to investigate the effect of 3 inch and larger pipe on 2 inch pipe targets at 1300 psi and 550oF.
In 1981-1984, the CEA-CEN (France) conducted tests and studies on the effects of pipe-on-pipe impact using 4 inch whipping pipe on 4 inch and 2 inch pipe targets. The testing program also included pipe whip impact against steel plates and concrete slabs. The test conditions were those of a PWR at 2400 psi and 600oF.
In 1983, Combustion Engineering performs tests on energy absorbing stainless steel honeycomb material.
In 1984-1987, the Pacific Northwest Laboratory conducted for the NRC a series of tests were 6 inch pipes were catapulted against 3 inch to 12 inch pipe targets, to investigate the potential for propagation of breaks by pipe-to-pipe impact.
In 1985 the CEGB and Magnox Electric plc (UK) conducted pipe whip tests on cantilevered pipe to determine the influence of several factors on the zone of influence swept by the whipping pipe. Factors investigated included the direction of the thrust force at the broken end, strain-hardening of the pipe material, whipping pipe with multiple bends, the difference in behavior between opening and closing bends.
In 1986, in an EPRI-sponsored study, the experimental results for two-phase jets from the Marviken tests were compared to numerical simulation (EPRI-NP-4362).
In 1988 a revised ANS 58.2 is issued, replacing the 1980 issue, introducing several changes including the use of the leak-before-break method.
In 1990, Siemens conducted tests to investigate what happens at the hinge section of a whipping pipe. The hinge section was also tested with a circumferential crack, to investigate the highly unlikely case of a pipe break occurring in a pipe that had a pre-existing crack at the buckle section.
In the mid-1990’s the University of Manchester (UK) conducted experimental and numerical studies of the plastic behavior, ovalization of the cross-section, and buckling of whipping pipes as a function of D/t (the ratio of their diameter to their thickness). The flow restriction caused by ovalization is also addressed in the study.