The Fatal First Natural Gas Well in Southern Arkansas

At 2:10 pm on Thursday, April 22, 1920, the small town of El Dorado, in Union County, Arkansas was busily going about its day. Over a dozen miles northwest, a drilling rig had just finished drilling the first productive oil well in Arkansas six days prior, the Hunter No. 1. However, just two miles^[1] southwest of town was another well being drilled that was going to make headlines as the first productive gas well in southern Arkansas. 

Top left: land owners, Mr. and Mrs. Frank Hill. Center: the Hill No. 1 well blowing water, gas, and mud into the air. Bottom right: Constantine Oil and Refining geologist, J. J. Victor.[2]

The Hill No. 1 well (Sec. 1, T.18S., R.16W.) operated by Constantine Oil and Refining Company of Tulsa, Oklahoma drilled to 2,226 feet, when it became an "unexpected 'gusher'" and a "great 'gasser'"^[3]. The Monroe News-Star newspaper of Monroe, Louisiana reported that it received a kick (i.e. anomalous influx of fluids), which can be extremely dangerous. The rig crew safely retreated to a distance as gas, brackish water, and mud spurted to nearly 130 feet into the air and the jet engine-like roar of gas "could be heard several miles"^[3]. This spectacle attracted "hundreds of people" to see a well reportedly flowing with 60 Mmcf (million cubic feet) to 100 Mmcf of gas per day^[3] although conservative estimates put it around 10 Mmcf^[2]. No equipment was available to control the pressure or to safely remove the nearly 1000 feet of drill pipe remaining in the borehole, but some was ordered that night from Shreveport, Louisiana, letting the well to flow uncontrolled for a few days. Leaving a release of flammable gas unchecked is undoubtedly very dangerous, but little could be done except to cordon off the area and post signs forbidding smoking^[4]. Those on the wellsite were cautious and wary of igniting the gas, but "those in charge of operations do not fear such a calamity"^[3]. Little could be done by the state, because at this time, Arkansas did not have a regulatory commission overseeing the oil and gas industry, unlike Louisiana, Oklahoma, and Texas.

Five days after the well blewout, gas and water continued to gush out of the well uncontrollably, although equipment had just arrived to control the well, but it would take some time to set up^[5,6,7,8]. Within the first few days of the blowout, the elevators on the derrick failed, possibly due to damage from the sand and fluids bursting out of the well causing the remaining drill pipe left in the borehole to fall to the bottom of the hole^[2]. However, investors were pleased to see a slight increase in the amount of oil^[5], which was more profitable than gas and the worthless 5,000 bbls (barrels) of brackish water produced per day. Constantine Oil and Refining Co. was working on plans to drill a second well nearby, but competition was closing in. Speculation of more oil and gas in Union County ran rampant, causing property leases for wells to jump drastically. Land within a mile-and-a-half was being sold for about $85/acre (appx. $1,070 in 2019)^[9], with one 40-acre tract selling for $4,500 (appx. $56,780 in 2019)^[6]. At this point, it was estimated that 50 Mmcf to 60 Mmcf of gas, water, oil, and mud was erupting out of the well per day, wasting about $3,000 (appx. $36,800 in 2019) of gas per day^[6,7].

Twelve days after the blowout, May 4, local visitors were still arriving to see the sight. The well’s roar could “be heard a half mile away”^[8].  But, investors were more interested in the amount of oil caught in the roots of plants along a nearby stream suggesting the possibility of more oil to come^[8]. Some said the gas pressure had reduced at this point, but the geologist, J. J. Victor, who had constantly been on the wellsite could see no difference. Although he admittedly stated that he had “become so accustomed to the well’s flow that he can see little difference”^[8]. Plans to cap the well were drawn up, as well as “two or three contingencies”, but the company was reluctant to cap the well in hopes that “any day to have oil in paying quantities”^[8].

The next day, other companies were flocking to Union County rigging up derricks and signing about 75 leases a day^[10]. Reports of former Arkansas state geologist turned professor at Stanford University, John C. Branner, was touring Union County collecting notes of the area^[10]. At this time, the Arkansas Geological Survey was inactive, but was restored the following year with John C. Branner’s son, George C. Branner, as the state geologist. Although the Hill No. 1 well was flowing 10 to 30 bbls per day, serious concerns of the stability and structure of the well was becoming public. Some worried that the “casing may be damaged” or “there may be a large cavern at the foot of the casing”^[10]. Concerns about not being able to “separate the gas and water” was also noted, and more direly, the “possibility of a crater, which oil men fear” from the collapse of the ground^[10].

Former Arkansas state geologist, John Casper Branner (image: Encyclopedia of Arkansas)

During the remainder of May, the well continued to be troublesome, refusing to be tamed. Equipment was being damaged and needed repairs, and attempts to cap the well failed^[11,12]. On May 24, a reduction in gas and water and an increase in oil brought some hope to Constantine for a profitable well^[13]. While the rig crew became hopeful with the arrival of a much needed valve from Homer, Louisiana allowing the crew to gain control of the well for the first time, one month after drilling ceased. Relief washed over Constantine and the rig crew after a loss of over $100,000 in wasted gas (appx. $1.2 million in 2019)^[12].

The hopeful situation once again changed for the worse on June 6, nearly six-and-a-half weeks after drilling ceased, when a “large crater developed…about a quarter mile from the Constantine [Hill No. 1 well]”^[1]. It was reported that this crater was about 25 feet in diameter with mud and sand gushing out of the ground like a geyser upwards of 30 feet into the air. The pressure subsided somewhat leaving a geyser of “[m]ud, sand, water and gas shoot[ing] up from 5 to 10 feet”^[1]. This was the beginning of further troubles for the well and for the local citizens. On June 8, the Monroe newspaper reported “more than 1,000 miniature gas wells…forming hourly…spread all over a radius of about a quarter mile of the Constantine [Hill No. 1 well]”^[1]. Local residents began “moving their household goods” in fear that their property could be damaged especially since that there was a “general belief that a huge lake will be formed as a result of the many imitators of the Constantine [Hill No. 1 well]”^[1]. Nearly 1,000 more “miniature” gas wells formed over the next day with the largest growing to “nearly 30 feet in diameter”^[14]. One of the largest of these erupted on June 12 “completely block[ing] one of the roads leading from El Dorado south”^[15].

Two of the craters creating fountains of water, gas, and mud near the Hill No. 1 well.^[16]

On the afternoon of June 13, almost two months after the well’s blowout, a group of locals at one of the largest craters a half-mile from the Hill No. 1 well arrived to photograph the phenomenon^[17]. One of the onlookers, a 35-year-old named Fred Hayes, pulled out a cigar and struck a match^[15]. That was all that it took for the geyser of mud and water to instantly became a towering inferno. The flame went “as high as seventy-five feet to 100 feet and may be seen for miles” and “appears to be getting larger constantly”^[17]. A trestle on the El Dorado and Western Railroad was nearly 30 feet away from the flaming fountain and the railroad began considering closing off the tracks due to threat from the fire. About ten of the onlookers, men and women, were badly burned and taken to the hospital in El Dorado, including an eight-year-old girl.

Following these eruptions, the Hill No. 1 well “has ceased flowing water and only gas is coming from it now”^[15]. Although good news for the well, the surrounding area was still having ongoing problems. On June 15, the fire at the nearby geyser was ongoing and three of those injured had succumbed to their injuries; the first being Mildred Martin, the eight-year-old. Although the flame was not spreading, gas from the ground kept feeding fuel to the fire leaving personnel helpless with the thought that it would be impossible to put the fire out^[17]. Discussions on what to do with the well in case the fire spread included the possibility of handing control of the well over to the government. Nearly resigned to hope that the well will burn or snuff itself out from mud, plans were being drawn to use a technique used in Louisiana to control oilfield conflagrations. This method involves bringing in locomotives on the railroad track that has been closed off to try to smother the fire with steam^[17]. However, it would take about seven to ten days “before preparations are completed”^[17]. During which time other “miniature wells [were] forming constantly” and first crater that formed has grown to about 60 feet in diameter^[17]. Visitors wishing to see the blaze were kept back by personnel acting as guards who also were enforcing a ban on smoking within two miles of the Hill No. 1 well^[17].

On June 17, reports circulated that a fourth victim of the fire had died from burns sustained on June 13 and the fire was continuing, although contained^{[18, 19]}. Nothing had yet been done to put it out, still in hopes that it will burn out naturally^[19]. On June 22, the news reported that a fifth victim succumbed to injuries, but the inferno ceased that morning ^[20]. After burning for nearly ten days, the Monroe newspaper simply suggested that it was likely due to a lack of gas flow into the geyser^[20].

After the wild blaze was extinguished, some sense of normalcy slowly began to return to the El Dorado area. Unfortunately, that lasted a month before interrupted by more trouble. A few days before August 21, a suspected leak in the casing of the Hill No. 1 well had to be mended. “When the leak was stopped the pressure…blew off the cap” of the well^[20]. Control of the well was restored shortly and the geysers “subsided and the gas pressure is practically under control”^[21].

The well remained stable and production was successful. Constantine Oil and Refining didn’t get profitable quantities of oil from the Hill No. 1 well, but the gas was able to be sold and they could claim the well as the first natural gas well in southern Arkansas. Contracts to supply natural gas with local power plants and two El Dorado high schools were signed. Pipelines would soon after be laid into the ground connecting the well. A second well would be drilled nearby by Constantine without detriment, followed by additional wells by Constantine and other operators^[22].

But, how did the geysers erupt around the well like they did and why were there craters? I'll try to explain, but I'm going to get into a bit of speculation here, since I don't know the local geology very well. First off, let's think about the blowout at the well. The fluids (gas, water, and oil) are trapped in pore spaces within sandstone rock (you can think of it as a water-saturated rock), a couple of thousand feet underground with nowhere to go.

A simplified cutaway of a poorly-cemented and poorly-consolidated sandstone with sand grains (brown) and pore-filled water (blue). 

The fluids and the sand grains are both under a lot of pressure. Again, there is a couple of thousand feet of rock pushing down on the reservoir, which is obviously a lot of weight (literally tons). The sand grains are pushing on the fluids and the fluids are pushing on the sand grains. When the well is drilled into the sandstone, there is suddenly a conduit to the surface, where there is waaay less pressure. So, the fluids escape up the borehole. As they rapidly do so, fractures may have developed linking the reservoir to the surface in the vicinity of the well. The fractures growth may have been aided by reorganization of the poorly-consolidated sand grains in the rocks. The sandstone that the well drilled into was of Cretaceous age, and may not have been well-cemented (well-consolidated). After they were rapidly drained of the bulk of the fluid (as well as some unconsolidated sand and sediment), the rock kind of deflated in a sense. The sand grains compressed together, collapsing the rock like a sinkhole and bringing down the overlying rock beds as well.

Drain the water and the sand grains compact into a tighter, closely spaced arrangement.

This isn't something that occurs in oilfields today, although ground subsidence is a very real thing as a result of oil, gas, and water wells, but it occurs gradually over a period of years or decades as fluid is removed in a controlled manner. In the case of the Hill No. 1, large amounts of fluid was released suddenly and under high pressure in a rock conducive to fracturing. I have never heard of this phenomenon before until I heard about the Norphlet Crater, which was a rather large hole (about 400 feet wide!) that formed from another well near El Dorado in 1922 which I hope to write about soon. 

Sources:

1, “Tiny gas wells by hundreds in Union County, Arkansas, start up around Constantine's well,” The Monroe News-Star, Monroe, Louisiana, 8 Jun. 1920, p. 1, print.

2, “Gas wells in Arkansas are turning lease tide that way,” The Oil Weekly, Houston, Texas, 8 May 1920, v. 67, n. 6, p. 21, print.

3, “Great "gasser" is brought in near Eldorado,” The Monroe News-Star, Monroe, Louisiana, 24 Apr. 1920, p. 1, print.

4, “Company geologist reports oil find in Union County “gusher” and gas pressure gets weaker,” The Monroe News-Star, Monroe, Louisiana, 26 Apr. 1920, p. 1, print.

5, “Oil traces at Constantine are said to be increasing,” Hunter planning to drill more wells, The Monroe News-Star, Monroe, Louisiana, 27 Apr. 1920, p. 1, print.

6, “Constantine gas well likely to be capped within next few days,” The Monroe News-Star, Monroe, Louisiana, 29 Apr. 1920, p. 1, print.

7, “Constantine well will be capped today,” The Monroe News-Star, Monroe, Louisiana, 30 Apr. 1920, p. 1, print.

8, “Graphic description of geyser of oil or gas at Constantine,” The Monroe News-Star, Monroe, Louisiana, 4 May 1920, p. 1, print.

9, “Hunter well reaming complet'd and casing to be placed today,” Constantine equipment arrives, The Monroe News-Star, Monroe, Louisiana, 28 Apr. 1920, p. 1, print.

10, “Constantine's well is still untamed thing,” The Monroe News-Star, Monroe, Louisiana, 5 May 1920, p. 1 & 6, print.

11, “Constantine well still spouts thousands of dollars worth of gas daily, flows wild for weeks,” The Monroe News-Star, Monroe, Louisiana, 18 May 1920, p. 1, print.

12, “Constantine well running wild for weeks, finally controlled,” The Monroe News-Star, Monroe, Louisiana, 5 Jun. 1920, p. 1, print.

13, “Constantine well gradually is turning to oil, gas decreasing,” The Monroe News-Star, Monroe, Louisiana, 24 May 1920, p. 1, print.

14, “More wells spring up,” The Monroe News-Star, Monroe, Louisiana, 9 Jun. 1920, p. 1, print.

15, “Gas fields of Arkansas afire many are hurt,” The Monroe News-Star, Monroe, Louisiana, 14 Jun. 1920, p. 1, print.

16, The Oil Weekly, Houston, Texas, 26 Jun. 1920, v. 67, n. 13, p. 32, print.

17, “Constantine area in danger of conversion into great volcano,” The Monroe News-Star, Monroe, Louisiana, 15 Jun. 1920, p. 1, print.

18, “Fourth victim of tragedy,” The Monroe News-Star, Monroe, Louisiana, 17 Jun. 1920, p. 1, print.

19, “Fred Hayes still serious,” The Monroe News-Star, Monroe, Louisiana, 22 Jun. 1920, p. 1, print.

20, “Constantine cap is blown off and old Vesuvius spouts again,” The Monroe News-Star, Monroe, Louisiana, 21 Aug. 1920, p. 1, print.

21, “Constantine "wonder well" to supply El Dorado with its gas,” The Monroe News-Star, Monroe, Louisiana, 30 Aug. 1920, p. 3, print.

22, “Geyser in Arkansas,” Oil, Paint and Drug Reporter, New York City, New York, 5 Jul. 1920, v. 98, n. 1, p. 19, print.