2009Innovating While Drilling®November/DecemberSafety and ESG

Modified BIBO, LOT applied to trip out from deep HPHT well with simultaneous loss, ballooning

Figure 1: A wellbore schematic of the subject well, a very deep HPHT well that had experienced simultaneous loss and ballooning. It had already reached its planned TD. The objective was to trip out for logging or other well operations while maintaining static hole condition.
Figure 1: A wellbore schematic of the subject well, a very deep HPHT well that had experienced simultaneous loss and ballooning. It had already reached its planned TD. The objective was to trip out for logging or other well operations while maintaining static hole condition.

By Fred Ng, Wild Well Control

The leak off test (LOT) and barrel in – barrel out (BIBO) procedures were modified and applied successfully to trip a very deep high-pressure, high-temperature (HPHT) well that was experiencing simultaneous loss and gain conditions. This is a common challenge for many HPHT wells, where margins between fracture gradient, mud weight and pore pressure are typically close. This article describes how these modified LOT and BIBO procedures were applied, as well as some of the risks and mitigations involved.

BACKGROUND

The subject is a very deep HPHT well that experienced apparent simultaneous loss and ballooning. Repeated application of lost-circulation material (LCM) pills and squeezes failed to solve the problem, as additional fractures seemed to develop whenever the well was re-pressurized by any significant circulating pressure.

Attempts to pull out of the hole (POH) resulted in apparent swabbing. A total of more than 2,000 bbl of mud had already been lost, which likely charged up a loss zone at total depth (TD), as well as other zones of induced fractures. The well was showing classic symptoms of taking and giving back mud at the same mud weight.

The mud system was synthetic-based mud (SBM), with a surface mud weight (SMW) of over 18.0 ppg. The well had reached its planned TD, and the immediate objective was to trip out for logging or other further well operations while maintaining static hole condition.

A wellbore schematic is shown in Figure 1. Although the concepts and occurrences described in this article are based on an actual case, well depths and some pressures have been modified for proprietary reasons. As a result, some of the mathematical details may not necessarily be consistent, but the technical and operational principles involved remain valid.

Figure 2: There should be a linear trend for pump pressure (PP) and volume of mud pumped in (VI) until PP reaches PFG1. Continued pumping past this point would result in a pressure that causes fracture propagation (PFGP). Large losses may resume.
Figure 2: There should be a linear trend for pump pressure (PP) and volume of mud pumped in (VI) until PP reaches PFG1. Continued pumping past this point would result in a pressure that causes fracture propagation (PFGP). Large losses may resume.

PROGNOSIS

Since pulling the drillstring resulted in apparent swabbing, an alternative was to pump out of the hole. However, as noted above, more than 2,000 bbl of mud had been lost at this point. The charged-up loss zones were showing classic symptoms of taking and giving back mud at the same mud weight. As discussed in a previous paper by this author (see footnote), this condition can continue until most or all of the lost mud is flowed back, and the induced fractures can close up. It was therefore critical to limit additional mud loss, which could create worse complications for tripping and other well operations.

BIBO-LOT IN PRINCIPLE

The operational objective at this point was to establish a pressure for pumping out of the hole with minimum mud loss. A process for establishing this is similar to performing a leak off test. This can be performed either with the well open or BOP closed with choke full open, whichever would be best for accurate measurement of return volumes.

Figure 3: This pressure plot shows a straight line from 0 to 3.5 bbl pumped, then a steeper straight line from 3.5 to 4 bbl.
Figure 3: This pressure plot shows a straight line from 0 to 3.5 bbl pumped, then a steeper straight line from 3.5 to 4 bbl.

The concept is shown in Figure 2 and described in the following procedure:

  • Pump mud into drillstring at the slowest practical rate using the cement pumping unit.
  • Plot pump pressure (PP) vs volume pumped in (VI) as well as volume of mud returned (VO).
  • There should be a linear trend for PP and VI until PP reaches PFG1. This is the pressure at which fractures begin to open, but not necessarily propagate.
  • Continued pumping past this point would result in a pressure that causes fracture propagation (PFGP), and large losses may resume.
  • As long as PP is below PFG1, VO / VI should be 100%. Beyond this, it would be less than 100%.
  • Mud loss can be minimized for pumping out of the hole with BIBO, using minimum practical pump rate and limiting pump pressure to less than PFGP.

PRACTICAL PROCEDURE

Due to difference of compressibility between the mud and the formation, actual results from the rig varied somewhat from the idealized version in the above theory. In addition, rather than pumping mud into the formation, it had strengthened enough to provide fluid return at a very slow pump rate of 0.25 barrel per minute (BPM), resulting in a true BIBO operation.

Figure 4: This volume plot shows that full returns started at 4.5 bbl pumped at a point where pressure peaked out at 320 psi. This is the charge volume of mud needed to pressure up the wellbore.
Figure 4: This volume plot shows that full returns started at 4.5 bbl pumped at a point where pressure peaked out at 320 psi. This is the charge volume of mud needed to pressure up the wellbore.

The pressure and volume plots in Figures 3 and 4 show what took place when pumping into the drillstring with the bit on bottom at TD:

  • The pressure plot in Figure 3 shows a straight line from 0 to 3.5 bbl pumped, most likely due to the fairly high compressibility of the mud.
  • From 3.5 bbl to 4.5 bbl pumped, the pressure plot shows a steeper straight line.  This is likely due to lower compressibility of the formation and ballooning of the wellbore.
  • The volume plot in Figure 4 shows full returns started at 4.5 bbl pumped, at a point where pressure peaked out at 320 psi. This is the charge volume of mud needed to pressure up the wellbore.
  • Based on the above measurements, a decision was made to pump out of the hole with a 300-psi limit.
  • This pumping procedure was applied to each stand to build up the needed pressure while monitoring volumes in and out.
  • For each stand, 4.5 bbl were pumped or a pressure of 300 psi was established, whichever occurred first, before pulling off the slips.
  • The stand was then picked up off the slips and pulled slowly while maintaining 300 psi. Pulling speed was adjusted to match steel volume removal rate with pump in rate.
  • Net volume in is monitored to ensure it is the same as the volume of steel removed.

As the bit is pulled up the hole, the above conditions can be expected to change with the increasing amount of actively affected wellbore. In this case, the loss events and treatment had occurred over a period of days, during which the mostly static mud appeared to have packed off at least partially around the bottomhole assembly (BHA) due to temperature conditions.

Figures 5 (above) and 6 show that, with the bit at the 10 ¾-in. top of liner at 16,000 ft, peak pump-in pressure and charge volume are higher and larger respectively.  At 16,000 ft, the target pressure and volume were modified to pump 5.0 bbl or establish 450 psi for each stand, whichever occurred first. Circulation was established when the bit was tripped to about half-way out of the hole.
Figures 5 (above) and 6 show that, with the bit at the 10 ¾-in. top of liner at 16,000 ft, peak pump-in pressure and charge volume are higher and larger respectively. At 16,000 ft, the target pressure and volume were modified to pump 5.0 bbl or establish 450 psi for each stand, whichever occurred first. Circulation was established when the bit was tripped to about half-way out of the hole.

As the BHA moves up the hole, increasing mud and hole volumes below the bit resulted in changes in response between pressure and volume. Figures 5 and 6 show these responses with the bit at the 10 ¾-in. top of liner (TOL) at 16,000 ft:

  • Peak pump-in pressure and charge volume are higher and larger respectively.
  • At 16,000 ft, the target pressure and volume for the procedure was modified to pump 5.0 bbl or establish 450 psi for each stand, whichever occurred first.
  • If a BIBO-LOT plot is done for each stand or every few stands, the change can be made gradually. In practice, it was sufficient to check these parameters, say, every 10 stands. In any case, once the bit is above the TOL, this would have been the interval to check and see if circulation can be established.
  • Circulation was eventually established when the bit was tripped to about half-way out of the hole.

RISKS AND MITIGATIONS

Coordination: While the principles involved are not complicated, the operation described requires very close coordination of the entire crew and service technicians, including the driller, mud logger, pumping unit operator, etc. The choreography involved can be especially challenging for an offshore rig or in the absence of day light, which makes it difficult to establish sight communications and references.

Figure 6
Figure 6

Returns in annulus: In this case, the formation had strengthened enough to provide fluid return at 0.25 BPM. This resulted in a true BIBO operation and provided VO/VI data to help establish pressure and volume criteria for the operation. If the annulus is packed off and no return can be observed, the procedure should be modified based on the charge volume needed to establish peak pressure, such as can be determined in Figures 2 and 4.

Pressure and volume measurement: Experience indicates that pressure and volume measurements from various sources on the rig can be different. Some of this may be due to the device used or location of the measuring points (e.g., standpipe gauge, pumping unit chart readings, mud logger readout, trip tank reading, fluid hold up in flow line or MGS system).

It is not uncommon to find false readings from a defective gauge. These parameters should be checked for consistency and any differences reconciled before starting the pumping procedure.

Flow back: When the pump is shut off, the well may flow back. This is expected since the 4- to 5-bbl charge volume has to flow back when pumping stops and the system relaxes to its uncharged state. This flow can be through the annulus if it is not packed off, or through the bit and drillstring if there is no float or the float malfunctions, or not at all if the annulus is packed off and the float holds.

CONCLUSIONS

Simultaneous loss and ballooning is not uncommon in HPHT wells. A modified BIBO-LOT process was applied to trip out from a very deep HPHT well that was experiencing these conditions. Successful application of this procedure helped to avoid further large mud losses in the charged-up zones and induced fractures, thus avoiding a more complicated and expensive scenario for this well. These same principles should be applicable to similar challenging drilling scenarios.

This article is based on a presentation at the IADC Well Control Asia Pacific 2009 Conference & Exhibition, 18-19 November, Bangkok, Thailand.

Fred Ng is general manager of engineering at Wild Well Control Inc. He has over 25 years of experience in operational, technical and management responsibilities. He served extensive assignments covering land and offshore drilling operations in the Gulf Coast, Alaska, Texas, Indonesia, Malaysia, China, New Zealand and Ghana. He is a mechanical engineer by training and holds a B.S. from the University of New South Wales and M.S. and Ph.D. from Texas A&M. He has taught in mechanical engineering and petroleum engineering at the University of Houston.

He is the author of “Kick Handling with Losses in an HPHT Environment,” World Oil, March 2009.

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button