The following technical abstracts are available from Cambridge Petroleum Software. Please contact us if you would like to request the complete article;
Stacking velocities or velocity functions in a field development? We can draw some significant conclusions by making simple comparisons between the two methods.
V0K or Mid Point Depth functions? How do they compare, and under what circumstances will they give different results? Which is the most reliable?
RayTrace 3D & IMRII or PreSDM?
How do we optimise the use of deviated wells data in velocity modelling?
Can I depth convert fault planes and fault sticks in VelocityManager?
During a drilling program of multiple deviated wells how can I use VelocityManager help me to actively manage in ‘real time’ my seismic interpretation and well prognoses?
Velocity Functions Too few or too many wells? –the dilemma.
Can I depth convert ‘overhanging’ Zechstein diapirs in VelocityManager?
Can the coefficients for the function Vo K, instantaneous velocity vs. depth, be computed accurately by optimisation?
How can I achieve effective ahead of drill bit forecasting with VelocityManager?
Residuals – Noise or data? How are residuals treated in VelocityManager?
HybridFunctions™: A useful approach to accurate velocity modelling with RMS data.
How do we optimise the use of deviated wells data in velocity modelling?
Optimised Residuals: A robust approach for the
computation of the residual surface.
Stacking velocities or velocity functions in a field development? We can draw some significant conclusions by making simple comparisons between the two methods.
There is often a temptation to use all available velocity data in a development environment with the expectation that ‘we can squeeze out every last bit of velocity information available’, but eventually we will have to make a judgement about which approach to velocity modelling is likely to give the most accurate results, velocity functions or stacking velocity.
It can be a very complex task, detailing and comparing all areas of uncertainty with both stacking velocities and velocity functions, for particular data sets. However, it can be informative to get an overview of the problems and appreciate the intrinsic limitations in accuracy associated with each method by studying the following two models...
V0K or Mid Point Depth functions? How do they compare, and under what circumstances will they give different results? Which is the most reliable?
Both functions V 0K (or Instantaneous Velocity) and Mid Point Depth endeavour to predict interval velocity as a function of depth of burial, honouring the physical principle that increased depth of burial goes hand in hand with increased density and velocity. The assumptions and mathematics behind each function are different, as are their limitations...
RayTrace 3D & IMRII or PreSDM?
For most geophysicists, there has been no alternative to Pre Stack Depth Migration (PreSDM) as the potential solution to solve both the migration and depth conversion problems in the presence of structural complication or steeply dipping horizons. This decision is not without penalty. It can be very expensive, impart significant delays on a project, and the cost may not justify the results. Too commonly the depth migration does not offer the anticipated improvement in seismic quality, as it retains some nasty depth artefacts beneath complex overburden; and the depth conversion accuracy, as judged by the quality of subsequent well ties can leave serious doubts about its overall accuracy, and its ability to give the definitive map for those ‘ all too sensitive low relief closures’. However there is an effective alternative...
How do we optimise the use of deviated wells data in velocity modelling?
Effective management of data from deviated wells is essential for accurate depth conversion. Failure to meet exact standards, can lead to inaccurate results  which will come back to haunt us. VelocityManager is specifically designed to model deviated well data, and is now widely regarded as the leading software in this field.
Projects involving deviated wells raise a lot of questions, because of velocity anomalies, but in most cases these can be resolved quite easily...
Can I depth convert fault planes and fault sticks in VelocityManager?
Effective depth conversion of fault planes requires accurate modelling of the velocity data together with careful integration of the fault grid data (or sticks). Failure to meet exact standards, can lead to inaccurate results  which will come back to haunt us. VelocityManager is specifically designed to model fault plane data in exact accordance with the velocity model derived for the velocity layers, and is now widely regarded as the leading software in this field...
During a drilling program of multiple deviated wells how can I use VelocityManager help me to actively manage in ‘real time’ my seismic interpretation and well prognoses?
In Effect, There may be two problems one needs to address from after prognosis
1) Display the engineers programmed deviated well track onto the seismic for QC purposes.
2) Monitor drilling results on the seismic and update the velocity model, or the interpretation in the event of any surprises with view to modifying the program whilst drilling.
Commonly the Geophysicist is draw to depth converting the seismic volume in order to display the prognosed deviated well, and monitor progress, but there are many advantages in continuing to work the seismic in time, and VelocityManager functionality has a vital role to play at this point...
Velocity Functions Too few or too many wells? – The dilemma.
The following investigation is based on well data collected from across multiple Quads for 9 layers, in order to evaluate the consistency and usefulness of this amount of ‘unedited’ data from across such an area with significant structural variations. Well ‘Tops’ data was taken from approximately 140 released wells, though not all horizons were present in all wells. All techniques used in this study can be equally be applied to deviated and horizontal wells for all velocity functions.
Our assumption is that if velocity function of time depth or thickness is to be reliable, it must be seen to predict the well data; and from a significant number of wells. To fulfil this criterion, we select the most appropriate function for each layer, and by derivation via optimisation, the coefficients must largely predict the well data, such that the residuals are comparatively small, and can be seen in map view to show a ‘smooth’ gentle variation which is consistent with either a small component of lateral variation due to lateral variation in rock properties, or an area of uplift, or both...
Can I depth convert ‘overhanging’ Zechstein diapirs in VelocityManager?
Effective depth conversion of salt diapirs requires not only accurate modelling of the velocity data, but also a specific ‘scheme’ for depth converting the complexities such as overhanging salt flanks which give rise to the problematic multiple time values for specific shot points. Failure to meet exact standards can lead to inaccurate results  which will come back to haunt us. VelocityManager is specifically designed to offer a number of features which can significantly help both in improving accuracy, and also by improving the ease with which the task can be achieved...
Can the coefficients for the function Vo K, instantaneous velocity vs. depth, be computed accurately by optimisation?
The Vo K function is the cornerstone function for predicting the interval velocity of clastics, and some other lithologies. Many geophysicists have, for convenience, presumed to approximate this function by the use of other functions such as ‘Interval Velocity vs. Mid Point Depth’ which can result in particularly unpredictable results, particularly for thicker intervals, and varying interval thicknesses.
In the past the Vo K function has attracted some criticism particularly where the coefficients have been measured from velocity logs. In so doing, it is intrinsically assumed that the increase in velocity from top to base of an interval will approximate to the same increase in velocity that is found with increase in depth of burial of the unit as a whole. In fact this assumption can lead to significant errors, particularly in areas of inversion, or where there is vertical lithologic variability, and thus velocity variations within the interval. This process requires that the coefficients Vo and K be corrected in what is often a very subjective manner.
VelocityManager uses an innovative and much more sophisticated and reliable process to calculate the values Vo K. The process, optimisation** computes a ‘global’ solution from the entire well data base which is empirical in the sense that it measures directly the values for Vo and K which are required to predict the interval velocity values encountered in all of the wells in the study...
How can I achieve effective ahead of drill bit forecasting with VelocityManager?
The progress of deviated and horizontal wells whilst drilling is commonly viewed and QC’d on a depth converted seismic volume. This can be useful unless problems are encountered with the interpretation or velocity model, necessitating a complete revision to the velocity modelling and depth conversion, in order to update the well prognosis ahead of the drill bit.  and also to redepth convert the seismic volume in order that monitoring can proceed. But these results may be required urgently!
The results can now be achieved very rapidly with VelocityManager.
The geophysicist may be drawn to depth converting the seismic volume in order to display the prognosed deviated well and monitor drilling progress, but in fact there are greater advantages in continuing to work the seismic in time. Working with depth seismic would necessitate redepth conversion of the seismic volume following any revisions to the interpretation or velocity model to ensure that the correct relationship between the well prognosis, ahead of drill bit, and the seismic is maintained. Working with time seismic only requires that the well prognosis is retime converted following any ‘surprises’ and changes to the model. This is now a viable option with VelocityManager, utilising its capability to:
a) Accurately convert well path data from depth to time.
b) Support very rapid revisions to the well prognosis ahead of drill bit....
Residuals – Noise or data? How are residuals treated in VelocityManager?
One of the most hotly debated aspects of function modelling is the treatment of residuals, and what they represent. To some “they are an unmanageable error which one tries to lose”. This view is often based on many years of experience in modelling within the software constraints of incorrect or inappropriate functions, where the residuals may well be chaotic. However, there is now compelling evidence to show that where modelling is implemented using a rigorous approach, with the most appropriate function, the residuals can form a coherent and essential element in the model...
HybridFunctions™: A useful approach to accurate velocity modelling with RMS data.
VelocityManagerII now offers a new approach to depth conversion from seismic processing RMS velocities, which allow us to achieve a significantly higher degree of accuracy in depth conversion with HybridFunctions™  even in areas of complex structure. HybridFunctions™ allows us to take advantage of the characteristics of functions which have a theoretically high degree of accuracy to spatial resolution, and apply these same functions, selectively, to RMS velocity data. The resultant velocity map can potentially have a resolving power some two orders of magnitude better than that of RMS data processed in a conventional manner.
How do we optimise the use of deviated wells data in velocity modelling?
VelocityManager is specifically designed to model deviated well data, and is now widely regarded as the leading software in this field. Accurate management of data from deviated wells is essential for accurate depth conversion. Failure to meet exact standards can lead to inaccurate results  which will come back to haunt us.
Optimised Residuals: A robust approach for the
computation of the residual surface.
The significance of residuals is an issue which achieves great attention in VelocityManagerII. Careful attention to the selection of the function type, function coefficients, and residuals can both enhance accuracy in depth conversion, and equally importantly, can clarify sources of error in the input data. VelocityManagerII now offers a new approach to residuals surface computation with optimised residuals which allow us to achieve a significantly higher degree of accuracy in depth conversion  even in areas of complex structure. Optimised residuals will overcome complex issues in selecting the most appropriate option in the computation of the residual surface. This is a very significant consideration, and can have a very significant effect in areas of structural complexity, or simply in areas of significant variation in interval thickness.
