Issue 104: Sprinkler System Seismic Research Moves to the Laboratory
By Russell P. Fleming, P.E., FSFPE
At a time when “resilience” has become a popular new goal, the earthquake protection provisions of NFPA 13 have received praise as an example of foresight and preparation. It should be recognized that the current provisions within the NFPA 13 standard are largely the result of taking advantage of lessons from actual earthquakes, most notably the 1989 Loma Prieta earthquake in northern California and the 1994 Northridge earthquake in southern California. Following both of those earthquakes, SFPE joined the National Fire Sprinkler Association and the National Fire Protection Association to co-sponsor public hearings to document experience for the NFPA 13 task group charged with improving the seismic readiness of sprinkler systems. In essence, the earthquakes served as real-life laboratories.
In recent years, there has been a shift to actual laboratory testing employing large shake tables to learn more about improving the seismic resilience of fire sprinkler and other nonstructural building systems, funded largely by the National Science Foundation. The motivation is that for almost all building occupancies, at least 70% of the original building cost is invested in the nonstructure. Nonstructural building features are usually more fragile than the structure due to the fact that the systems were individually developed in the absence of seismic concerns, and then pushed together in a “nonstructural sandwich” above suspended ceilings where they can damage each other through their interaction under earthquake forces.
Many in the fire protection engineering community are aware of one recent research program that was completed in 2013 involving the fire protection engineering department at Worcester Polytechnic Institute and the outdoor shake table at the University of California San Diego.1 The program involved 13 motion tests of a 5-story reinforced concrete test building, including the Maximum Considered Earthquake (MCE). Following shaking, 6 fire tests were conducted on the 3rd floor that included a pressurized wet pipe sprinkler system. The basic observation was favorable to the NFPA 13 provisions: “Sprinkler systems were not damaged from seismic tests and worked well in the fire tests.”
The fire protection engineering community is not as familiar with the recent 5-year program of the Network for Earthquake Engineering Simulation (NEES) that received “grand challenge” funding through the National Science Foundation and involved multiple institutions. Starting in 2008, the goal of the program was to develop a simulation capability for ceiling-piping-partition nonstructural system interaction. A comprehensive experimental program that would provide the basis for mathematical simulation involved test facilities at the University at Buffalo, the E-Defense Facility in Japan, and the University of Nevada – Reno.
The program at the University of Buffalo2 examined ceiling-piping-partition interactions, and included two test series related to fire sprinkler systems. The first involved testing of 48 tee-joint combinations using pipe from ¾-inch to 6-inch in diameter, and variations of steel pipe with threaded joints, steel with grooved connections, and CPVC with cemented joints. The joints were tested under reverse cyclic loading to determine the rotational capacities at which leakage and/or fracture occurred.
In the second test series, two-story arrangements of the three types of materials and joints were dynamically tested on a shake table both with and without NFPA 13 bracing. All three fully braced specimens performed well under a Maximum Considered Earthquake (MCE) level of loading. The unbraced systems suffered damage to sprinklers, failures of vertical hangers, and a branch line fracture.In the second test series, two-story arrangements of the three types of materials and joints were dynamically tested on a shake table both with and without NFPA 13 bracing. All three fully braced specimens performed well under a Maximum Considered Earthquake (MCE) level of loading. The unbraced systems suffered damage to sprinklers, failures of vertical hangers, and a branch line fracture.
Following the tests, models were introduced to simulate the nonlinear moment-rotation behavior of tee joint components made of the various materials and joint types. The results from numerical simulations showed close agreement with the experimental results in terms of displacement, acceleration, and moment – rotation relation at piping joints.
In the Japanese tests,3 fire sprinkler systems were installed on the 4th and 5th floors of a 5-story building constructed on a large shake table. The sprinkler systems were identical, with steel pipe connected with flexible couplings on the mains and threaded joints on the branch lines. Hangers and restraints were provided in accordance with NFPA 13, but variations were used in the sprinkler/ceiling interface, including use of a 2-inch oversized gap or flexible drops. Flexible drops utilize “flexible sprinkler hose fittings”, first allowed by NFPA 13 in the 2007 edition, to connect the branch lines to the sprinklers in a suspended ceiling. The main difference between the 4th and 5th stories was that the suspended ceiling was itself braced on the 5th floor, but not on the 4th floor.
Major observations included limited benefit from lateral bracing if the system was subjected to strong vertical forces. Also, due to the twisting moment generated from armover drops, long branch line pipes with several unsupported armover drops were found to twist around the branch line threaded connection points. The oversized gap configuration was not effective in preventing damage at the ceiling panel/sprinkler interface, but the use of flexible drops substantially reduced the piping-ceiling interaction. In some cases, the pounding of the pipe against the hanger ring under vertical forces led to failure of the hanger ring connection.
Back at the University of Nevada – Reno4 the effort was focused on pulling it all together, with strength testing conducted on individual hangers, braces and restraints, and the seismic vulnerability of the system determined by combining the effects of various components. Three-dimensional modeling was also performed on a sample hospital fire sprinkler system, with 96 tri-axial floor acceleration histories. The end product of the research program is the first-ever analytical modeling methodology for fire sprinkler system piping, with fragility curves developed at both the component and system level.
Some of the findings of the fragility study are not directly relatable to NFPA 13, mainly because it is not important if some components are stronger than others provided the rules of the standard limit the loads on all components to acceptable values based on material strengths or product listings. But some of the findings confirm long-standing views of the NFPA 13 experts. For example, it was found that armovers and drops to ceilings are the most vulnerable points within the systems. As might be expected, system failure probabilities were less for dry pipe systems than for wet systems. And the use of flexible couplings on system risers, as required by NFPA 13, works well to prevent leakage due to interstory drifts.
In the future, the modeling capability achieved by the grand challenge will serve as a new tool for continued refinement of the NFPA 13 earthquake protection rules. In the short term, the confirmation of both the vulnerability of drops and armovers, and the advantages accompanying the use of flexible drops, could lead to a push for the use of flexible drops where sprinklers serve suspended ceilings in high risk earthquake areas.
Russell P. Fleming is with the National Fire Sprinkler Association.
References:
- Brian Meacham, Ph.D., P.E., FSFPE, Haejun Park, Ph.D. and Jin-Kyung Kim, Earthquake and Post-Earthquake Fire Performance, Fire Protection Engineering, Issue 82, March 2014.
- Yuan Tian, Andre Filiatrault and Gilberto Mosqueda, Experimental Seismic Study of Pressurized Fire Sprinkler Piping Subsystems.
- Siavash Soroushian, Keri L. Ryan, ManosMaragakis, Eiji Sato, Tomohiro Sasaki, Taichiro Okazaki, Lee Tedesco, Arash E. Zaghi, Gilberto Mosqueda, and Dennis Alvarez, Seismic Response of Ceiling/Sprinkler Piping Nonstructural Systems in NEES TIPS/NEES Nonstructural/NIED Collaborative Tests on a Full Scale 5-Story Building.
- S. Soroushian, E. M. Maragakis, A. E. Zaghi, A. Echevarria, Y. Tian, and A. Filiatrault,Comprehensive Analytical Seismic Fragility of Fire Sprinkler Piping Systems, MCEER Technical Report (MCEER-14-0002), August 2014.