Depth repeatability of gamma logs
- 2 minutes read - 345 wordsHi All,
Let’s touch on one of the many subtle issues around wireline logging and gamma logging in particular and it’s flow on effects. I have seen a few arguments on where to place the stratigraphic contact on a gamma log: above the peak, on the peak, below the peak, this then rapidly escalates to we need to get all the surfaces to snap to interpretations exactly!
A solution for the problem might be as simple as “within the assay interval”.
However I thought that it would be helpful to have repeatability of the depth measurement of gamma log demonstrated which you can then bring data to the how stringent am I going to be argument.
To generate a quick estimate I hunted through the WAMEX database to extract data from a project that contained enough in-rod and openhole gamma logs to calculate a sensible histogram, which we have here. You can hopefully see that there is a bias between the IR (In-Rod gamma) depth and the OH(Open Hole), this is due to many reasons but in this case I would argue that the biggest issue is the difficulty in getting an estimate of ground level when the drill rig is in the way.
The standard deviation of the offset between measurements is around 0.15m the and the mean is effectively 0m, which is pretty good.
![histogram of offset between open hole and in rod gamma](/images/OpenHoleVsIr_hu90106f158b64e0f4607afa7686f7a1d9_26868_640x480_resize_q95_h2_box_3.webp)
A diagram on cross correlation for those who are not familiar: cross correlation is the process where we calculate a correlation between each signal by sliding the two signals along one another like in the gif below, the point where the correlation is maximised is assumed to be the offset, which in the case of this data is 0.
To finish up, I have a couple points:
- The correct place to place the put the contact is on the shoulder of the peak half way between the measurement spacing.
- There is a good chance that the first RC sample is longer than the nominal sample interval due to the interaction of the starter rod and drill table.
Thanks,
Ben.