Sewer Sinkhole Study Using TEM
Society of Exploration Geophysicists, The Leading Edge, March 2006.
Authors – Norman R. Carlson*, Scott A. Urquhart*, Zonge Engineering and Research Organization, Tucson, Arizona, U.S.A.
Article – [pdf] INFR_LeadingEdge-SinkholePaper
Slide presentation – [pdf] sewer intregrity survey TEM system
On 7 September 2002, two large sinkholes opened suddenly on a major, five-lane roadway in Tucson, Arizona. The exact cause is still controversial, but the sinkholes were apparently related to failures near 90° bends in a 35-year-old, 6-m deep, 1.1-m diameter sewer line known as the outfall interceptor. A third, smaller sinkhole developed just a few city blocks away on the same night at another 90° bend in the same sewer interceptor line, this time at the intersection of two residential streets. With the sudden overnight development of three sinkholes on city streets, the potential for harm to the public was obvious. Had the sudden subsidence under the five-lane road occurred during rush-hour traffic, for example, injuries could have ranged from minor to fatal.
The City of Tucson contracted Zonge to assist in rapidly evaluating other locations along the 8km length of the sewer interceptor line to locate other problematic sites before they developed into costly, possibly dangerous, sinkholes. The areas around 90° bends were to be surveyed first, followed by a survey of the entire line.
The conditions of the survey path eliminated several geophysical methods that we might normally use to detect subsurface voids or unusual fluid accumulations. Since the majority of the sewer line is beneath paved roads, galvanic methods such as resistivity and IP, which require good electrical
contact with the ground, were considered too slow and difficult. The low-resistivity soil also limited the use of ground-penetrating radar, since the depth of investigation would not have been sufficient for the goals of the survey. Although transient electromagnetics (TEM) would not normally be our first choice for void detection, the change in resistivity from an unusual fluid accumulation should be detectable with TEM, and the system can be cart-mounted for rapid coverage; thus, we elected to test TEM first over the third, small sinkhole to determine if it appeared anomalous relative to the surrounding, presumably good, sections of the same sewer line. This sinkhole area was weakly anomalous: lower resistivity relative to nearby sections of the sewer, probably as a result of the fluid escaping from the corroded and collapsed sewer.
We proceeded to survey seven additional sites near sharp bends and eventually the entire 8km line using TEM. The geophysical survey clearly showed anomalous areas later confirmed as compromised sections of the sewer interceptor line, and may have averted a catastrophic failure involving a natural gas pipeline found suspended in the void above a 5-meter area of dissolved sewer pipeline and only 1 meter below the roadway—a discovery made by the county after further investigation of anomalous areas.