Dear Sandile,

velocity analysis is one of the main and more tricky step performed during the processing of active seismic data. By means of it, the velocity information is estimated from the seismic traces.

It can be divided into 3 categories: time velocity analysis, depth velocity analysis, well velocity analysis. Each of them has its own characteristics and field of application: the first two pertain to data acquired at the surface, the third to data acquired into wells.

As you know seismic data are amplitude values of the movement of the ground (it could be displacement (rarely), velocity (often) or acceleration) as a function of time and space. By knowing the position of each source and receiver it is possible to predict the position of the signal of the seismic waves on the traces. But to do this you need to know the velocity, which is usually unknown. So, if some simple velocity model (the most simple is the homogeneous one) is assumed, the equations for the travel times of the wave can be written and, because you can recognize the signals on the traces, you can derive an estimate of the velocity. By knowing the velocity you can go on with the following step of the seismic processing (stack, migration). Obviously the procedure in the real world is much more complex and it requires powerful programs to do it, but the principle it that. The first to be performed during processing is the time velocity analysis, because you derive a velocity function/model that is function of time, like the seismic traces. Such velocity is used to perform the stack of the seismic traces and the time migration. At the end of this process you get the seismic image of subsurface (seismic line) as a function of time. If you are working in an area with high structural complexity, the time image is not completely reliable or clear and so it is necessary to get it in depth. The most correct procedure to do it is to apply depth migration in which the used velocity model is in depth. To derive it, the depth velocity analysis must be performed. It is applied after the time processing and it is computer and human intensive.

Well velocity analysis measures directly the velocity data of the subsurface by acquiring the data with the receivers into the well while the source can be either into the well or at the surface. The first is the case of the log velocity data, recorded at ultra-sonic frequencies, which give the velocity values very close to the well (few inches). Instead, the source at the surface is a standard seismic source and the record the arrivals of the seismic waves in the proximity of the well (some tenths of meters). It is called Vertical Seismic Profiles and it gives a direct conversion table between time and depth.

Well seismic ties relates to the procedure to match the data recorded into the well on the seismic data acquired at the surface (seismic lines). Well data are in depth while seismic data can be in time or depth and the passage between the two is through the velocity. Even if they are in depth the agreement with well data is not always straightforward because velocities values are obtained in very different condition, as described above. Moreover, consider that in the well vertical velocity is measured while on the surface horizontal velocity is estimated and they don’t match if there is anisotropy.

Two classical books on the item:

Ozdogan Yilmaz, Seismic Data Processing, Society of Exploration Geophysics, 2001.

Bob Hardage, Vertical Seismic Profiles, Geophysical Press 1983 (don’t know if still around).

I hope not to have made too much confusion.

Best regards,

Lorenzo

Velocity Analysis is the process of calculating the stacking & migration velocities of the Earth layers by using different offset reflected Seismic wavelet for a single location called CMP.

These velocity analysis lead to compute the Vrms which it lead to compute the Vinterval , and Vaverage for the consecutive layers of the Earth layers.

Seismic well introduce Vaverage & Vinterval of the well location.

So, the correct velocity analysis ( Vaverage & Vinterval ) of the Seismic wavelet should ties with the corresponding one of the Seismic well

In general, we calibrate our velocity analysis of the Seismic data with the Seismic well velocity.

Thank you all so much for the input you have given, i now have an understanding towards the topic. Thank you everyone!!!

It is better if u study it from Seismic Data Processing  written by YILMAZ.....

The stack velocity is an approximation of the Vrms, because it is derived from a truncation of a time serie, and can be used only in the data processing. There are several problems when the offset is large against the reflection time. Errors can derive when the signal/noise ratio is poor, from dipping reflecting interfaces, uncorrect processing procedures, ... At large depths and small offsets the NMO curvature becomes ininfluent and the computations fail.

The interval velocity is derived from the stack velocity with the Dix eauation and is affected by the above errors and approximations.

The well velocity corresponds to the interval velocity but with much higher resolution and the errors can be reduced.

The match between the two velocities is usually difficult or impossible.

Vertical velocities are given by checkshots ans VSP's in near vertical wells/ short offset surface source. ( tiny velocity differences are sometaimes observed versus frequency... still an R&D subject).

Long range surface seismic refraction yield the horizontal velocities of a diectrete series of refraction markers corresponding to interfaces with substantial velocity increment;, ans the averahe velocity from surface to refraction marker

In reflection seismic, Vrms or stacking velocities ar about 3-5% higher than vertical velocities.