As drilling engineers, we are often involved in optimising the drill string design to improve tensile, torsional and hydraulic performance. However, all too often we leave drill-pipe inspection to the QA/QC department. In doing so, are we increasing the risk of failure on our wells?
Drill String Component
The concept of drill-string design is well understood: torque and tensile loads can be modelled accurately for a range of load cases that are expected within the given well operation, using torque and drag software. This allows components to be specified such that the applied load remains below the rated load, with some reserve capacity built in for use in an emergency (e.g. overpull if the string becomes stuck). This process is called overload design and is so well understood that failures in the overload criteria represent <20% of all drill string failures. The concept of overload design goes hand-in-hand with drill string inspection and allows the engineer to set acceptance criteria on various drill-string attributes such that the rated load of the drill-string is within the design calculations. The specification of an inspection class has often been sufficient for the design engineer to outline the acceptance criteria, with premium class specifying that at least 80% of the new drill-pipe wall thickness remains, with the subsequent downrating of tensile and torsional ratings. While this may be acceptable for most routine, or mid-range drilling conditions, it may not be so applicable to ERD wells, where bulk loads often approach the rated loads of the drill string. In this instance, the cost of failure warrants additional inspection to reduce the probability of failure.
Drill string inspection
Once inspection criteria have been specified, and the pipe suitably inspected, the drilling project can proceed. The design engineer is then faced with the problem of when to schedule a re-inspection: can the drillstring be used for multiple wells or should the drill-pipe be re-inspected every well? In terms of overload design, the answer is simple: drill-pipe dimensions can be checked at surface to ensure they remain within the inspection class, allowing the actual loads to remain below the tensile rating. However, the situation becomes much more complicated when fatigue loads are considered. In the context of drilling, fatigue is the progressive and irreversible damage inflicted on the drill-pipe as it is rotated within a curve: one side of the pipe it is bent into tension, with the other side in compression, while the process reverses and repeats for every revolution of the drill-string. This cyclic process causes failure of a drill string at levels well below the yield stress of the material.
Image of fatigue failure
The fatigue life of a Drill String Component (figure 1) has two phases, crack initiation and crack propagation. The initiation phase represents the vast majority of the components life and is typically defined by the components S-N Curve. This represents the number of cycles (revolutions) that can be completed before failure may occur as a function of applied stress. A typical S-N curve for a G-grade DP is shown below. This defines the bulk stress amplitude verses the number of cycles to failure. The deviation from a straight line represents the fatigue endurance limit, whereby if applied stress remains below this level, the pipe could experience this level of stress for an infinite time.
However fatigue failures are still observed, even at stresses levels below the fatigue endurance limit – to understand why this may be the case we need to understand that these curves are generated under laboratory conditions and do no adequately represent an active drillstring component which has numerous stress concentrators; internal / external upsets, slip cuts and pits which have accumulated over years of service. Secondly, the S-N curve is highly dependant upon the conditions of service where typical S-N curves are represented in air, the presence of corrosive materials present in a wellbore (H2S, CO2, Chlorides) will rapidly accelerate the fatigue failure due to reduction in the fatigue endurance limit. Without knowledge of how the string was used and in what environment or accumulated fatigue to date), it becomes impossible to accurately predict the failure of a given component using a S-N curves.
The alternative approach to managing fatigue life is to look at the crack propagation phase in more detail – this assumes a crack is present and calculates whether the service loads that the pipe will experience will cause that crack to grow to the point of failure. There are still some unknowns in the crack growth model (based on accumulated fatigue), however these are assumed constant in the analysis which enables the designer to select the best alternative (drill string configuration) to help maximise fatigue life. The analysis therefore allows the designer to manage the fatigue risk, but does not ensure that fatigue failure will not occur. The key challenge thereafter is establishing the correct inspection frequency to identify a crack before it leads to a problem.
It is typical for operators to outline an inspection frequency based upon a pre-determined number of rotating hours (e.g. 1500hrs), or footage drilled, however this may not be applicable to critical wells where fatigue failures may become more common and / or the cost of failure is high. How then does the design engineer determine the appropriate inspection frequency? One option is the define the inspection frequency based upon the drilling conditions being observed. With this it is important to understand that drill string fatigue is not uniform – higher fatigue rates will l be observed across a build section that pipe which has largely been used in a tangent section. The crack growth model does allow for analysis of this by splitting the drill sting into sections, the cumulative damage that the pipe has seen based on given drilling parameters, dogleg severity and tensil loads can be estimated:
With this sort of analysis, the drill pipe can be managed / rotated to ensure that fatigue life is spread more evenly across the drill string and through time a threshold damage point number can be established based on field experience which would trigger reinspection. The assessment of the cost of failure can help set the threshold figure where in general, for critical wells inspection should be conducted more frequently.
The subsequent inspection will indicate whether a crack exists and segregate this, however it will not capture instances where the fatigue life is approaching the fatigue endurance limit, therefore if is recommended to perform Brinell hardness / Charpy VN testing on tube joints to see if the pipe deviates significantly from new pipe as this will give an indication of induced fatigue and may help tailor the inspection frequency if drill pipe history is known.
Experience has shown us that fatigue is one of the primary causes of drill sting failure – with the current economics, the cost of failure has never been so high, therefore improving drilling performance will required us, the drilling engineer to better understand drill string fatigue and work together with QA/QC departments to help manage drill sting failure.