For example;
When considering Mn removal efficiency of a particular filter, removal efficiency can be compared using actual influent and effluent values.
Similarly, removal efficiency can be compared using log values of both influent and effluent.
Are there any significant benefits of doing that?
using log values is fairly common practice in for example petroleum engineering when for instance cross-plotting porosity vs permeability.
porosity values of a reservoir rock typically range between 5% and 35% of bulk volume. Permeability values commonly range between 0.001 Darcy and 10 Darcy.
Cross-plotting porosity values vs log values for permeability than typically gives a useable correlation coefficient whereas plotting porosity vs absolute permeability values does not result in a useable coefficient
W.J.E. Van de Graaff Thank you very much for sharing your thoughts. I'll check on instance cross-plotting porosity vs permeability graphs.
In the Oil Exploration industry , its an important practice in many aspects. We have logs that measure many paramters in oil wells. and equivalent lab measurements are designed to measure the same parameters on actual picked samples (core data). this for instance is used in two main aspects. first to assess the uncertainty in logs and design better logging tools to reach minumum difference between log/lab measurements. second is to calibrate log data to real sample data which leads to a better prediction of reservoir distribution away from drilled wells when implementing actual values into models and when integrating with other related measurements in the 3D (Seismic data).
In oil and gas exploration, log values are extremely important in calibrating log response to different lithologies. Taking actual cores is very time consuming and consequently expensive, especially in deepwater wells. So, if you can calibrate your log response to actual rock samples, it makes your trust in log responses where you do not have rock samples more reasonable.
Benefits of transforming parameters by a function can relate to the identification of dependencies between pairs of parameters. Functions used might be or not uniquely inversible and be applicable to all values or not. Dependencies can be measured by various means, including correlation coefficients. A drawback can be the fact that dependencies between parameters can differ, hence transforming a parameter can bias toward some correlation and against others, or more generally, generate spurious correlations. This observation underlies the saying in applied physics that "with a log / log plot I can correlate anything". In my domain of activity, it is typical to correlate porosity to logarithm of permeability. There are valid reason to so so (too long to explain here, and not your question).
Mahmoud Noaman & Sharon Cornelius Thank you very much for sharing your valuable insights here.
Supun,
please bear in mind that in the various answers two very different meanings of the term ‘log values’ are used.
In my response I used ‘log value’ in the mathematical sense of a logarithmic value of a parameter: in my case permeability values which in real rocks can range from 0.000001 Darcy to 10 Darcy or more. Expressing these values on a logarithmic scale helps in establishing relationships with porosity.
The other meaning of ‘log values’ that is used in some other responses is ‘downhole measurement with a geophysical logging tool. For example a Gamma Ray log which measures natural radioactivity around a borehole in the subsurface.
- Badges
- Science topic