March 2000 Vol. 221 No. 3  Exploration

Perry A. Fischer,  Engineering Editor

Bad science? Just dial 1-800-TREMBLE

All geologists and geophysicists studied earthquakes at some point in their educations. Worldwide, each year averages about 18 earthquakes of Richter magnitude (M) 7.0 or larger, 120 of M 6.0–6.9, 800 of M 5.0–5.9 and an estimated 6,200 in the M 4.0–4.9 range. Because of their devastating potential, much time and money has been spent trying to predict earthquakes – and thus save lives.

In the quest for predictors, seismologists and others have measured changes in seismic activity, volcanic gasses, fault displacement and groundwater (level, temperature, chemical composition and production rates), just to name a few. Even oil wells may show changes. Thus far, nothing has worked reliably.

However, there have been a few successful earthquake predictions. The most famous was that of the 1975 Haicheng earthquake in northern China, where an evacuation warning was issued the day before an M 7.3 earthquake. Local people reported the early warning signs; they were safely outdoors, in cold February weather, when the earthquake struck.

In the months preceding the event, widespread reports of changes in land elevation, groundwater levels, peculiar animal behavior and many foreshocks led to a lower-level warning. An increase in foreshock activity prompted the evacuation warning. However, even in areas where foreshocks are common, there is no way to distinguish a foreshock from an isolated event. In America’s Pacific Northwest, there is no evidence of foreshock activity for most earthquakes. In San Francisco in 1989 and Kobe in 1995, the quakes came without warning.

A much-less-known success story occurred before the 1976 Tangshan quake which killed at least 250,000; but in Qinglong County, just 60 mi away, officials heeded warning signs and persuaded much of the population to camp out. Although 18,000 buildings collapsed when the quake hit, no one was killed.

Using a set of assumptions about the rate of stress accumulation and fault mechanics, the USGS made a prediction about Parkfield, California: an M 6.0 earthquake would strike between 1988 and 1992. The quake never happened, but an M 6.0 earthquake is still predicted for Parkfield. If it happens, a dense network of instrumentation will determine whether precursors exist and provide new details on fault-rupture mechanics.

USGS and Stanford Research Institute funded a 1-800 number in 1977. The focus then was on animal behavior, and people would call someone who would write down their reason (dog barking, geese quacking, etc.) for quake prediction. Some of those who worked on the project (there were up to 1,800 volunteers) said that after four years, just when there was beginning to be clear correlation between several earthquakes and the number of calls – the project was de-funded. At the very least, the project was labor-intensive, difficult to analyze and sounded silly.

It seems to this editor that a well-publicized 1-800 number could automatically record call frequency and perhaps other information, which, over time, might provide correlative evidence of an impending earthquake. People could call for whatever reason – whenever. It would seem to be an impossible system to corrupt. The data might not reveal anything, but no matter what the number or frequency of "crank" calls, if there is shown to be a predictive correlation between call frequency and earthquakes, it would be worth the relatively low cost.

Perhaps scientific dogma prevents such a system. Such dogma is good and necessary – up to a point. It helps prevent wasted time and money, and preserves public trust in what otherwise might be failed, too-far-out endeavors. To be scientific, earthquake predictions should state when and where the event will occur, as well as why the prediction was made. It is the latter requirement that simply counting calls and area codes does not provide.

Of course, one could complicate the process with a long telephone menu that begins, "Press 1 if your water well just turned into a boiling geyser, press 2 if your cat is stuck to the ceiling, press 3 if you were informed by a time-traveling alien from a galaxy far, far away . . ." and so on.

So, do your oil wells act strangely before a quake? Is your goldfish doing back flips? If you live in an earthquake-prone area, you may want to see a 1-800-PREDICT telephone line funded. It’s relatively cheap and, even if it’s bad science, it just might save lives.

San Joaquin Deep. Interest in the southern San Joaquin Valley has been high since the spectacular, high-rate blowout of Bellevue-1 on Nov. 23, 1998. A relief well had to be drilled to kill it, but it disproved the conventional belief that the deep Tremblor sands were too tight to produce.

The Berkley Cal Canal-1 well, drilled to 18,100-ft TD, penetrated 1,200 ft of the Tremblor formation, resulting in 775 ft of net sand. Well completion began on January 20, and an initial 10-ft zone was perforated in the Lower McDonald. Hydrocarbon flows were non-commercial.

Several additional zones in the Tremblor, McDonald and Upper McDonald still need to be tested. The seven-company consortium will defer further deepening or completion efforts until reservoir quality and other data are gathered from ongoing drilling efforts in the San Joaquin. The partners’ next San Joaquin well should spud by third quarter, 2000.

Berkley is also part of a nine-member consortium exploring in the East Lost Hills area in the San Jaoquin. Its Bellevue-1-17R well was tested in December and flowed between 1.3 and 5.0 MMcfgd. Pressure-buildup analysis showed that only the uppermost sand contributed to the flow. Since the production liner is uncemented, the company is reviewing its options carefully.

Also drilling ahead is the partner’s East Lost Hills Berkley-1 well. The well was at 18,300 ft in early February, and into the Tremblor, with good gas shows and about 720 ft of net sand encountered in four units. Planned TD is 20,000 ft. Two additional East Lost Hills wells are scheduled to spud by midyear.

Tri-Valley Oil and Gas Co., as operator in a 10-company consortium, spudded EKHO-1 on February 7 this year. The well is also targeting the deep Tremblor at 19,500 ft.

Comments? Write: fischerp@gulfpub.com

Copyright © 2000 World Oil
Copyright © 2000 Gulf Publishing Company