A common concern voiced by users of the ISCWSA error model is that the results that when they compare error ellipse values from different directional drilling software packages, they get different answers. In general, the immediate reaction is that one package must be wrong. However, the ISCWSA models leave a number of decisions up to either the user or software implementer and these should be taken into account when comparing results. See reference [7] for a fuller discussion on this topic.
What confidence limits are being reported?
The models make no comment on what confidence limit (number of standard deviations, sigma level) results should be given too. This is typically a user defined option in the software and values of 1, 2, 2.79 and 3 are all commonly used in the industry.
What co-ordinate system are ellipse dimensions reported in?
An error report may simply specify semi-major axes or semi-minor axis without detailing which co-ordinate system is being consider. Typically, the results can be in North-East-Vertical or Highside-Lateral-Alonghole axis. Some packages can output in only one or the other axis systems, other packages allow the user to select which ones are used.
What value is being reported?
Typically, a package will output semi-major axis and semi-minor axis. In some cases, (particularly TVD spread) it is 2 x semi-minor axis that is output (semi-diameter?).
What revision of the error model is implemented?
As detailed in the MWD section there have been a number of successive revisions of the MWD model and not all software packages will be at the same revision level.
What error magnitudes are being used in the tool model?
In general, the ISCWSA does not standardise or define the tool models to be used. It is only the mathematics that is standardised and it is left up to the survey contractor to supply appropriate tool models. The only exception is the ISCWSA has standardised on tool models for standard MWD and MWD with Axial Interference corrections. In other cases, the tool model comes supplied with directional package or can be edited by the user. Sadly, it is not uncommon to come across survey databases in use with tool models containing a very reduced set of error sources (or equivalently all the magnitudes set to zero.)
Feedback
Post your comment on this topic.