Home Page

Introduction | Recognition | Causes and Types | Significance | Control Methods | Alternate Kill Methods | Conclusion | Next Article |

Part 6 - Underground blowouts. Underground blowouts between subsurface intervals are common and can result in a significant escalation threat if not recognized quickly and controlled correctly

 Brian A. Tarr, Mobil E&P Services, Inc., Dallas 
 L. Flak, former Wright, Boots & Coots employee

This article provides basic information on (1) how to recognize an underground blowout and (2) methods that can be used to regain well control. 



Underground blowouts involve a significant downhole flow of formation fluids from a zone of higher pressure (the flowing zone) to one of lower pressure (the charged zone or loss zone.) They are the most common of all well control problems. This phenomenon differs from crossflow, which typically occurs within a long perforated interval and involves little or no reserves loss or escalation hazard.

Underground blowouts can occur in drilling wells or producing wells, and are thought to be the most common in the latter because of tubulars corrosion in older completions. Unfortunately, no statistics on underground blowouts are available because most go unrecognized or unreported. What is known is that many surface blowouts begin as underground blowouts. And prompt, correct reaction to an indicated underground flow can prevent an even more serious and costly surface blowout. That leads to the observation that "experience is the best of schoolmasters, only the school fees are heavy."l 



Fig. 12 shows how an underground blowout may occur in a drilling well. But field personnel commonly fail to recognize the symptoms-most often, they are focused on curing presumed loss circulation. The following events indicate underground flow:


Most underground blowouts that occur while drilling result from lack of sufficient kick tolerance. Kick tolerance is the kick intensity (amount of underbalance) that can be shut-in without exceeding the fracture pressure of the weakest exposed formation after taking a given volume kick.2

Lost returns occur when kick tolerance is exceeded. But sometimes lost returns may occur before the kick is taken and an underground blowout will result. An example is drilling with excessive mud weight and surging the hole on a trip in. Resulting lost returns drop fracture pressure from breakdown pressure to a lower fracture extension pressure. This induces a kick that starts underground flow to the fractured zone. Another cause might be losing returns to a depleted reservoir with high pressure permeability exposed elsewhere in open hole.

During drilling, casing holed by drill pipe wear or pipe defects can result in sudden lost circulation and an underground blowout. In producing wells, internal tubing corrosion or pipe defects can lead to failure and sudden imposition of tubing pressure on production casing. Defects or external corrosion of this outer casing can lead to either a subsurface or surface blowout depending on depth of the flowing zone.

Gas flow after cementing is a major cause of surface annular blowouts. Less recognized is that annular bridging or top-out cement jobs can divert gas flow underground. Surveys of many multi-well gas fields indicate some underground flow likely occurred after cementing. Natural formation bridging and scale deposition shut off most of this flow.

Many operators have been surprised when a temperature or noise log run in a shut-in well years after completion indicates crossflow. Generally, such flow is of little consequence. But there have been instances when this flow has led to surface broaching, fresh water aquifier contamination and shallow supercharging. Identification is difficult as temperature and noise logs must be run in a stable shut-in wellbore. 



Besides losing reserves to charged zones and possible environmental impacts (fresh water aquifer contamination, shallow supercharging), there have been cases when gas or water have flowed into partially depleted oil reservoirs. In one instance, a deep H2S gas reservoir in Iran flowed for over a year into a shallow sweet oil reservoir. Six nearby oil wells began producing nearly 100% H2S, which was flared to limit expansion of the gas migration front until flow could be stopped by a deep relief well.

Gas migration into a zone can be mapped using modern seismic techniques. This can assist greatly in locating a relief well or determining if flow is continuing after the well bridges or kill operations have blocked the surface flow path. Fig. 13 was constructed from seismic during a North Sea blowout. Note increasing migration of the gas with time. It should be pointed out that charging was occurring up dip from the wellbore. No gas is charging down dip of the wellbore.

Many surface blowouts through drill pipe from high temperature-high pressure wells are caused when a deep underground flow at FBHP lowers the drill pipe mud column. The mud column will fall until it equalizes with FBHP and slowly be replaced by blowout fluids if mud pumps are shutdown. As drill pipe pressure drops due to this fluid exchange, the borehole can become unstable and collapse around the bottomhole assembly This "bridges off" the flow and allows FBHP to build back to SIBHP. Drill pipe hydrostatic which is equal to FBHP thus is subjected to the higher SIBHP. If the bit is below the annular bridge and no check valve is in the drillstring, drill pipe pressure will increase rapidly. Fluid hammer can occur when mud in the drill pipe is pushed rapidly to surface by formation fluid entering through the bit. A surface blowout then occurs if rig pump pop valves open and/or surface valving fails (cuts out or cannot be closed against the flow).

This type of blowout will generally sustain if little surface solids production occurs. Fig. 14 shows flow discharge from a pop valve in just such a case. In this well, two kelly cocks and one standpipe gate valve were cut out by fluid hammer, leaks of gas and 17.9 ppg hematite oil mud.

Surface broaching of an underground blowout can lead to loss of rig and severe environmental impact. As surface access is unavailable after surface broaching, flow must be controlled by a relief well. 



Fig. 15 shows a top kill flowchart that outlines steps to be taken if an underground blowout occurs while drilling. This flowchart was developed by Mobil E&P Services' drilling technology group based on Wessel and Tarr's paper.2

Initial steps are to mobilize on location cementing pump(s), additional mud storage, and cement batch mixer(s) if available. Mix and store at least one additional hole volume of mud on location. While mixing mud, bullhead water down the annulus to the loss zone to minimize annulus surface pressure and prevent subjecting casing, wellhead and BOPs to gas. This will assist temperature log interpretation by defining a temperature gradient at the loss zone. Consider running a calibrated rate gyro to provide better relief well targeting. Fracture extension pressure can be estimated by adding surface pumping pressure to water hydrostatic to the loss zone. The top kill attempt consists of the following steps:

  1. Slow annulus pumping rate, continue annular water injection with cementing pump.
  2. Pump water or mud down drillstring at 90% of maximum possible rate using rig pumps until pressure stabilizes. Record stabilized pressure and rate. Increase pump rate to maximum and record stabilized pressureand rate.
  3. Stabilized drill pipe pressure is a function of the annulus two-phase (mud and gas or mud and oil) flow, hydrostatic and friction. If single-phase mud flow is attainable in the annulus, then the well is dynamically killed. Drill pipe pressure for single-phase flow can be accurately determined. If this pressure is achieved, the well is killed.
  4. Lost circulation materials can be added to the mud to obtain a static kill after pumps are shut down.
  5. If a dynamic kill with water or mud is not achieved, the recorded stabilized two-phase flow pressures developed during the attempted kill, in combination with results of the pressure/temperature log, can be accurately analyzed2,3 to determine what will be required.



If a top kill is impossible and normal hydrostatic control can't be restored, other procedures can be tried. A major complication to restoring normal circulation with a single weight uncontaminated fluid is possible supercharging.

Supercharging can make it impossible to achieve a normal hydrostatic kill without first isolating the flowing zone from the charged zone. This procedure is used as the primary kill method when a dynamic kill attempt cannot be completed or converted to a normal static kill without re-start of flow by the supercharged zone. Isolation can be accomplished by bridging (natural or induced), plugging (lost circulation materials, soft plugs), blocking (gunk, sodium silicate, cement) or mechanical means (openhole packers, cased hole packers).

Natural bridging controls most underground blowouts because exposed shales cannot withstand resulting pressure differentials. Natural bridging can sometimes be induced to shut off underground flow by reducing FBHP via surface venting. FBHP can be high if flowing zone permeability is high and high wellbore fracture pressure limits FBHP- which can be sufficient to support exposed shales. Surface venting at a high rate can drop FBHP below fracture pressure and cause bridging.

Plugging the flow path or the charged zone with lost circulation materials, barite plugs or diatomaceous earth squeezes is rarely effective in controlling significant underground flow, particularly if FBHP is equal to formation fracturing pressure. Tremendous quantities of drilling mud and lost circulation material have been pumped away in attempts to plug off a charged zone. Many times these are attempts to "regain circulation" when an underground blowout went unrecognized. Barite plugs are generally effective only if hydrostatic control is regained long enough to allow barite to settle. As with lost circulation material, use of barite plugs to control a severe underground blowout is generally a waste.

Blocking the flow path with reactant plugs of fast-setting cement, gunk or sodium silicate reacted with cement or C aCl2 brine can be effective. Gunk was first developed by Goins as a means for squeezing off a lost circulation zone. 4

Gunk generally consists of 150 ppb bentonite and 150 ppb cement mixed in diesel oil and reacted with a fresh water drilling mud. The reaction is nearly instantaneous at a 2:1 to 1:1 gunk-to-mud ratio. The firm "breaddough-like" mixture of hydrated bentonite and cement can be easily drilled, but will handle high differential pressures given sufficient plug length.

Alternately, 200 ppb guar gum and up to 100 ppb of fine lost circulation material can be used in diesel for a gunk plug that will react with saltwater flows, brines or salt muds. Guar gum can be replaced with modem high molecular weight polymers. Using these polymers with powdered CaCO3 as the lost circulation material produces a gunk that is 98%+ acid-soluble when reacted with brine.

Invert gunk consists of 275 ppb of amine clay (amine-treated bentonite) mixed in water and reacted with an oil flow or oil mud. Mixing a water mud as a primary kill fluid that contains 50 ppb amine clay is useful in killing oil flows. The amine treated clay will react when mixed with oil and markedly increase viscosity of the water-based kill fluid. This provides greater annular friction and helps limit mud losses after the well is killed.

Plugs of sodium silicate solution reacted with CaCl2 brine have also been used. Less recognized is that a mixture of 3:1 cement-to-sodium silicate will flash-set. Densified and highly retarded cement is generally used to obtain sufficient pumping time.

Any of these reactant plugs requires:

Reactant plug use to control severe underground blowouts should be done with the assistance of personnel experienced in their application. Misapplication can plug off the well above the underground flow and isolate the surface from the problem. This has generally been the case when cement was pumped without first controlling the flow.

Mechanical isolation using packers snubbed in below a hole in tubing or casing and set to isolate the hole has been done. Open-hole packers have been snubbed in and set near the flowing zone. Modern open-hole packers are available to handle high differential pressures (7,000 to 9,000 psi) if annular clearance between packer and hole is limited.

Coiled tubing has been used with these packers for control of underground flow in a few cases. A major difficulty is that once the packer is set and now is controlled, the situation is much like having a "bull by the horns" as hydrostatic control or bridging/plugging isolation is still needed 



Underground blowouts are a growing problem because of aging wells. Operators need to closely monitor existing producing wells for signs of problems. Most of the well control work done by author Flak last year involved producing wells-and tubulars corrosion was the single largest cause.

Operators many times fail to respond immediately and correctly when an underground blowout occurs. That makes control more difficult as flow paths erode, downhole tubulars degrade (erosion added to corrosion) and supercharging occurs. In a drilling well, early recognition is a problem because indications of underground flow are masked by operations to restore circulation. The list of well conditions provided herein can be used to determine if an underground blowout exists. The flowchart provides step-by-step instructions for implementing a top kill to control the blowout.

If a top kill is impossible, alternatives exist particularly if there are two independent flow paths to allow mixing of reactant plugs into the flow. This has been accomplished with coiled tubing, snubbing and directional relief wells. Mechanical plugs also have been used to isolate the flowing zone. Experienced personnel are required to simulate flow paths, make kill calculations and apply reactant plugs. 

Coming Next

Shallow Gas Blowouts.How to handle shallow gas blowouts.
Next Article

Literature cited

1. Carlyle, T. (1795 1881), Miscellaneous Essay, 1:137.
2. Wessel, M., and B. A. Tarr, "Underground well control: The key to drilling low-kick-tolerance wells safely and economically," SPE Drilling Engineering, December 1991, p. 250.
3. Smestad, P, 0. B. Rygg, and J. W. Wright, "Blowout control; Respose intervention and management, Part 5," World Oil, April 1994, pp. 75-80.
4. Goins, W. C., et al., "Method and composition for sealing lost circulation in well,," U.S. Patent 290016, June 27, 1961.   

The authors

Brian A. Tarr is a drilling advisor at the Mobil Exploration and Producing Technical Center, Dallas, Texas. His special interests include wild well control, directional drilling and surveying, and horizontal drilling. He holds a BS degree from Manchester University, England, and an ME degree in petroleum engineering from Heriot Watt University, Edinburgh. He is a Registered Petroleum Engineer in Texas.

L. Flak is a former Wright, Boots & Coots employee.