Smart flowback tool targets automation of kick, lost-circulation detection process
By Katie Mazerov, contributing editor
An automated flowback monitoring technology has been developed to detect well kicks and lost-circulation events earlier than conventional manual methods and to generate an alarm alerting drillers and drilling engineers to take appropriate action. A case study, SPE/IADC 163474, describing how the Smart Flowback Fingerprinting technology was developed and deployed in a field test, was presented by Sven M. Haberer, Baker Hughes, at the 2013 SPE/IADC Drilling Conference and Exhibition in Amsterdam this week.
“Macondo clearly demonstrated how kick can affect the safety of people and the environment,” Mr Haberer said. “But kick is also one of the major causes of nonproductive time (NPT).” Deepwater studies have shown that NPT related to kicks and lost-circulation events can amount to 4.5% of the total well construction time, he added. For a 90-day well at a spread rate of US $1 million/day, that would translate to an estimated $4 million.
Kicks frequently occur during drill pipe connections, and flowback fingerprinting, or monitoring changes in mud pit volume, has been the method most commonly used for more than a decade to detect such events. Less common detection methods are mud return flow (flow meters) and standpipe pressure.
“Currently, the industry uses a manual alarm system that relies heavily on the experience and intuition of the engineer performing this critical safety monitoring task,” Mr Haberer said. “And, as with any manual system, anything that can go wrong, will go wrong. The idea behind this new system is to make the kick detection more intuitive, self-adjusting and as automated as possible to remove the human bias.”
Removing the guesswork
The Smart Flowback Fingerprinting technology uses statistical analysis to interpret flowback data during connections, taking the guesswork out of this critical task by completely automating kick/loss detection. The system generates intelligent alarms based on current well conditions, so kicks can be detected up to one connection earlier than with conventional methods.
“The system ensures that flowbacks exceeding normal volumes are always detected in real time, without human intervention, and also minimizes false alarms,” Mr Haberer explained. “It also automatically generates threshold and alarm curves based on statistical analysis of previous flowback profiles.”
As part of the field test, the system was tested against both real-time surface data and archived databases from other wells that had experienced events, such as influxes and mud losses. This was done to refine the methodology to take into account such factors as flowback exclusion – instances where pumps were turned off briefly or operating at a low rate.
No kick or well control incidents were detected during the field trial. In order for the technology to be fully commercialized, a number of areas are being enhanced. These include the development of new algorithms to account for rig heave on the surface pits being measured to avoid false positives at the beginning of the flowback cycle. Tools for displaying real-time drilling data during remote collaboration also are being developed.
“As deepwater wells become more complex, pressure-related problems such as influxes, wellbore breathing or losses also will become more challenging,” Mr Haberer said. “We are confident the Smart Flowback Fingerprinting will be another tool in our toolbox of intelligent systems that will detect kicks early, prevent NPT and increase safety.”