It has long been acknowledged that compensator lock up during completions and well test operations could have catastrophic consequences. In the past and for many operators today this hazard is managed procedurally with additional maintenance checks and rigorous adherence to procedures; other operators are moving toward engineered solutions.
All floating drill rigs utilise some type of compensator to accommodate the movement of the rig with the rise and fall of the sea whilst the work string is fixed down hole, this is particularly important during completions and DST well test operations when the work string is also full of hydrocarbons. Compensators fall broadly into one of two general types, passive and active. Very simply a passive compensator uses air cushioned pistons which can compress and expand like a spring with rig motion, whilst an active system utilises a computer, with sensors that detect acceleration, which controls the main winch unit or draw-works accordingly to compensate for rig motion.
Like any complex system compensators are subject to various types of fault, mechanical, electrical, hydraulic or human error. There is one particular failure mode - lock up - which is of interest in circumstances where the workstring is "pinned to bottom" i.e. anchored at the sea bed. In this type of failure, the compensator fails rapidly and effectively becomes a rigid part of the rig. Should a lock up occur during a pinned-to-bottom operation and at a time when the rig is heaving more than say a meter or two, there is a significant risk such an event would result in catastrophic failure of the upper work string as the upward force generated by the buoyancy of the rig with the rising sea generates an excessive tensile force in the workstring which is fixed at the sea bed. The resultant reaction forces could propel the separated sections of the work string into the derrick and result in numerous significant dropped object hazards. At the same time pressurised fluids including hydrocarbons would be released either to the marine riser or at the drill floor creating a fire and or explosion hazard. The risk associated with the hazard outlined above is of course greater in shallow water where the pipe stretch available is minimal; however it would still exist in Deepwater if heave conditions were great enough. Additionally the Deepwater situation would have the added problem of a much greater hydrocarbon inventory in the landing string.
Whilst the statistics are scanty many operators in the industry acknowledge that the frequency of these type of failure events, based on the information to hand, is significant enough to at least indicate the likelihood on a risk matrix is greater than “rare”. IADC/SPE 59216 Unintentional Compensator Lockup Risks, Consequences and Measures presents three such events which occurred in the North sea during the 90's; from the reports in this paper we can readily conclude that this hazard has a high consequence potential, i.e. the events described in this paper could have resulted in multiple fatalities. Anecdotally when the authors discuss this issue with peers or during our well test courses we hear of stories where people have experienced or know of some kind of compensator event that has occurred elsewhere. I for instance have heard of an event some years ago where a flowhead fell over against the dolly track after the compensator failed during a well test, in this case no containment failure occurred. In June of 2011 the compensator locked up on a semi submersible in Australia during slickline operations on a completion resulting in a 300 K lb overpull on the landing string; luckily conditions were benign and the driller reacted quickly but the potential was there to part the landing string. We have attended drilling morning meetings where as an aside the rig reports “problems with the compensator last night”. During drilling operations compensator problems are largely inconsequential so the issue is not singled out for attention beyond any other minor mechanical failure that might occur, and so no serious effort has been made to catalogue these events.
Additional debate takes place around “Active versus Passive” compensator systems, some operators seem to have the perception this risk is singular to passive systems. In reality both systems have been involved with serious faults. Some very serious problems have occurred with active heave systems - just recently we have heard that a driller ran a software diagnostic on his console, which unbeknownst to him, resulted in the top drive automatically travelling the full distance to the rotary table. The reality is that regardless of which system your rig has and regardless of how well maintained it is, it is still a "single barrier"; some operators are now seeking to implement a second engineered level of protection against compensator failure.
One such solution involves the introduction of an in-line passive compensator to back up an existing active system. This solution is very costly and may not be suited to every rig due to the derrick interfaces and the engineering timeframe required. There are also some concerns and uncertainties about having dual compensators online at the same time. i.e. both the passive and active heave in operation simultaneously.
Perhaps a more practical solution is a set of shearable telescopic bail arms in place of the rigid bail arms normally used during a well test. Such units have been in use in Norway for some years, and though they have never been used “in anger” they do promise protection from over tension in the event of a compensator lock up.
An alternative and cost effective option exists to re-design the subsea test tree slick joint so that it allows the work string to strip through the pipe rams following a lock up event, thereby preventing an over tension failure. Designs for both of the above already exist.
The authors are not aware of other engineered solutions out there, though there may well be others. Details regarding some of the above examples are available as data sheets found at the links given below.
We would welcome any comments or experiences from readers regarding compensator failure experiences and other solutions.