Forensic Seismology: New Mexico Pipeline Explosion Seismic Study

On 19 August 2000 two seismometer networks in southeastern New Mexico recorded signals from a natural gas pipeline explosion. Analysis of the particle motion, arrival times, and durations of the seismic signals indicates that three impulsive events occurred with origin times of 11:26:18.8 1.9, 11:26:43.6 2.1, and 11:27:01.7 2.0 (UT). The first event was caused by the explosive blowout of the buried, high-pressure pipeline, and the second event was caused by the ignition of the vented natural gas. The nature of the third event is unclear; however, it was likely created by a secondary ignition. There were also two extended seismic events that originated at the same time as the first two impulsive events. The first resulted from the preignition venting of the gas and lasted for about 24 sec, while the second resulted from the postignition roaring of the flames and lasted for about 1 hr. Many of the source constraints provided by the seismic data were not available from any other investigative technique and thus were valuable to a diverse range of parties including the New Mexico state police, law firms involved in litigation related to the accident, the National Transportation and Safety Board, and the general public.

This research is published as: Koper, K., Wallace, T., Aster, R., Seismic Recordings of the Carlsbad, New Mexico, Pipeline Explosion of 19 August, 2000, Bull. Seism. Soc. Am., 93, 1427-1432, 2003.

The explosion created a large crater shown above (the State Police GPS location is (32.03785,-104.0286)). The crater is 86 feet long, 46 feet wide, and 20 feet deep. The people camped near the pipeline may have been unaware of its exact location because it was buried 15 feet underground, although the pipeline emerged from the ground to cross the Pecos river at a nearby bridge. The cause of the explosion was officially investigated by El Paso Natural Gas and the National Transportation Safety Board

Three groups of seismometers were deployed at the time of this explosion. Yellow triangles correspond to permanently deployed stations near Socorro, New Mexico; red triangles correspond to permanently deployed stations near Carlsbad and the Waste Isolation Pilot Plant (WIPP); blue triangles correspond to temporarily deployed stations that are part of a coincidentally deployed New Mexico Tech, seismic experiment funded by the National Science Foundation, RISTRA, to image deep earth structure. RISTRA is supported by the IRIS/PASSCAL Instrument Center at New Mexico Tech. The explosion is clearly visible on seismographs recorded by the permanently deployed Carlsbad stations (in red). We are currently analyzing data from the temporarily deployed stations (blue). The signal is very subtle or not visible at the Socorro stations. All clear signals are due to sound waves propagating in the atmosphere. The pressure variations in these sound waves cause elastic motion near the Earth's surface which is sensed by seismic instrumentation.

Locations of NM Tech/RISTRA stations that recorded seismic and/or air wave signals are

WIPP ANTR 32.2493 -103.405
WIPP CBET 32.4202 -103.989
WIPP CL2B 32.2642 -103.879
WIPP CL7 32.4128 -103.808
WIPP GDLM 32.2017 -104.48
WIPP HTMS 32.4725 -103.634

RISTRA TX06 31.967103 -103.706837
RISTRA NM07 32.084544 -103.839986
RISTRA NM09 32.326484 -104.118304
RISTRA NM10 32.472927 -104.267218

The three waveforms above show vertical short period (this instrumentation is primarily sensitive to frequencies above about 1 Hz) ground velocity at three of the permanent Carlsbad stations (in red). They are arranged according to distance from the explosion site, with the closest station (approximately 23 km distant) appearing first. The x axis shows time in seconds, and the y axis is proportional to ground velocity. The data are consistent with at least three significant separate explosions occurring within a time span of approximately 40 seconds, the first being much smaller than the subsequent two. Other smaller explosion events are also visible later in the recording. The approximate origin times for the first three events observed at station CL2B, estimated by Al Sanford and Kuo-wan Lin of New Mexico Tech using a sound velocity of approximately 335 m/s,

#1 05:26:19 MDT (Weak)
#2 05:26:43 MDT (Strong)
#3 05:27:02 MDT (Moderate)
Shutdown ~06:25 MDT

Under the simple assumption that the energy in each explosion is proportional to the square of the recorded ground amplitude, and normalizing by the amplitude and energy of the first event, we can crudely estimate the relative energy in each blast, from CL2B we have

#1 05:26:28 MDT Energy: 1
#2 05:26:52 MDT Energy: 11.8
#3 05:27:10 MDT Energy: 5.8

However, note that the energy received from event #3 at CBET is comparable to that received from event #2. This variation may reflect explosion directionality, and/or changing local atmospheric conditions (perhaps due to the heat pulse from the explosion) which can dramatically affect sound/seismic coupling.

This seismogram shows the blast recorded at the nearest permanent Carlsbad station. The time scale shown here is much longer than that of the figure above. An extended coda of seismic energy continues for nearly one hour after the explosion. This is associated with the roaring of the flames from the ruptured pipeline. The end of this coda is consistent with the time that the gas was shut off by the El Paso Natural Gas Company.