Gasoline in Ground Feeds Fire
A brush fire that first appeared to be like any other brush fire of rather limited area will long be remembered by men in the East Providence, R.I., Fire Department. The burning vegetation was extinguished routinely, but flames continued to flare up in piles of granite that had been dumped in a field on the east bank of the Seekonk River.
The smoky fire started on the morning of last April 18, probably as a result of boys using matches in the area. By the time the last flame went out about a month later, 35,391 gallons of gasoline had been pumped out of the ground and, it was estimated, more than 20,000 gallons had burned.
Two pipelines about a foot apart, one for gasoline and the other for No. 2 fuel oil, ran through the area from a tanker unloading dock to the Getty Oil Company and Humble (now Exxon) Oil & Refining Company tank farms about 2 1/2 miles away. The two companies operated the pipelines jointly.
Fire allowed to burn
After evaluating the situation, Chief Michael J. Fox decided it was more prudent to let the fire burn while a fire company stood by. Although there was a chemical warehouse several hundred feet from the fire area, there was no real exposure danger, whereas nonburning fumes that would have followed extinguishment would have caused an ignition danger over an undetermined area.
The two oil companies were told of the situation, but at that time there was no definite knowledge of what was burning. An air test was made on the two pipelines Friday, April 21. The pressure held on the 14-inch fuel oil pipeline, but the pressure fluctuated in the 12-inch gasoline pipeline. A trench was dug to expose more than 100 feet of pipeline near the fire area but nothing was found amiss.
In response to a request from Captain Robert D. Sauer, fire prevention officer of the East Providence Fire Department, Professor John O. Edwards of the Brown University chemistry department visited the site Friday morning to try to determine the fuel feeding the fire. Edwards talked with representatives of both Getty and Exxon and they decided that “some type of spill or leak of a petroleum fuel had occurred” and that the search for the source should be made immediately.
In his later report, Edwards said that “in some places the flames were colorless, indicating a volatile fuel with adequate air for complete combustion. Some of the flames were yellowish (indicative of a carbon fuel).” He also reported “many small fissures in the ground from which wisps of grayish smoke emanated. Occasionally one could hear a sound like that of a gas oven being lighted.”
Workmen dug a hole about 100 yards from the flames, on the Seekonk River bank, Saturday afternoon. The hole, about 16 inches in diameter and about 16 inches deep, quickly filled with “a blackish oily liquid immiscible with water.” A sample of this liquid was taken to the Brown University chemistry laboratory for gas chromatograph and infrared spectra tests. The sample appeared to be gasoline, but it had changed during the undetermined time it had been in the ground, Edwards stated.
Another sample, “showing a reddish tint superimposed on a brownish color,” was taken from a surface pool about 15 feet from the flames. A gas chromatograph test of this sample and nuclear magnetic resonance tests of both this sample and the previous sample identified the liquid in the ground as gasoline. Edwards said that the tests did not establish whether the gasoline was Getty or Exxon. Both companies used the same pipeline. However, the tests showed that the gasoline was partly or wholly high test.
Edwards’ report stated that the several ground samples taken and gasoline samples provided by the two oil companies showed many things in common. Edwards explained in his report that the presence of toluene, along with other aromatic hydrocarbons, and tertiary carbon-hydrogen bonds were important in the matchups of the ground samples with known gasoline samples. The reddish tint in two of the soil samples was believed caused by the dye used to identify premium gasolines. The tests made at Brown were confirmed by other tests made at the Exxon laboratories in Bayonne, N.J.
Meanwhile, a second air test of the pipelines was made Monday with results similar to the first, and further excavation was started.
The search for the pipeline leaks was spurred by the laboratory identification of the seepage as gasoline by Monday night and by the digging of holes in the area that, filled with what was later identified as gasoline and water. At first, a backhoe was used, but later a drag line was brought in for the job so that the machinery would be farther from the digging and gasoline vapors. The maximum depth of any hole was 20 feet.
The first leak in the gasoline pipeline was uncovered Wednesday about 100 feet from where the first pipeline excavation work had stopped. Eventually, six holes, ranging from about the diameter of a lead pencil to that of a man’s forefinger, were found in the 3/8-inch wall of the gasoline pipeline.
Air-operated pumps used
Pumps operated by compressed air were brought in to recover the gasoline from the ground. With the use of long air lines, the compressor motor was kept far from the vapor area as the pumps worked in the holes. Ground conditions resulted in wide differences in the seepage in the holes. In a test for flammable vapors in one hole, the flammable vapor meter needle remained on zero, but in a hole only 15 feet away, the needle went off the scale. By the time excavation work was completed, some 200 feet of pipeline had been exposed.
The area was kept under police guard for two months after the brush fire, until the area was considered to be safe. While the recovery of gasoline was going on, dry chemical extinguishers were kept at the site as a precaution against any flareup of fire on the surface of the seepage in the holes that were dug. However, the gasoline recovery work was completed without incident.
Installation of the gasoline pipeline was approved in 1941, and it was believed that it was laid shortly after World War II. When gasoline was flowing through this pipeline, a pressure of about 90 psi was used, it was reported. The fuel oil pipeline was believed to be about 20 years older than the gasoline line, although no records were readily available.