Cant help with the shale permeability.

But regarding choosing a design value based on a data set, I can say there is no one correct method. You can choose arithmetic average, geometric average, harmonic average. Then you can take the average of the averages, you can just adopt the lowest value, you can calculate a 95% confidence value based on statistics....shall I keep going.

This is where engineering judgement comes in to play. For your specific case of 1.5 or 1.6, I would pick the more conservative value since they are so close.

Thanks Eirechch, your message was a relief finding that I am not the only one being uncertain about it.

I guess by harmonic you meant K = L/(Ll/k1 + L2/k2), where L = L1+ L2. Then may be I take the average of arithmetic (parallel) and harmonic (series). I leave out the geometric mean as cannot yet be convinced why to employ it. In term of adopting the lowest values, will it be usually a conservative approach?

I dont even know how to calculate the harmonic mean. I just use excel and assume its right :)

Of course not, it depends, if youre doing infiltration calculations then 1.5 is more conservative, if you are doing dewatering then 1.6 is more conservative.

As with any shales, wouldn't the bedding planes and their attitude (horiztontal, vertical (as I had in Malaysia), fractures-joints and that) be one of the controlling factors in permeability? You are really looking for the coefficient of secondary (planes) permeability - not the matrix.

Thanks BigH for your point, which i understand that would be the so-called mass permeability compared to material permeability. There is no obvious fissures in the split spoon sample. Although the texture within the sample shows very fine fissures that do not seem to be open with naked eyes. A photo is attached if it may help.

literature survey suggested K in the range of 1E-9 to 1E-14 ft/sec for shales. Again uncertain what average value one can take as the arithmetic average always errs on the upper level.

• https://files.engineering.com/getfile.aspx?folder=759a5150-2b1f-4c5f-801e-7

If you can find a recording on youtube, watch Martin Preene's lecture 'Thinking Conceptually About Groundwater'. Not that it's Texas specific but he has some very succinct and experience based points about selecting "permeability" values. One thing to consider is: What permeability value would you get from a lab test? A cpt correlation? a single-borehole falling or rising head test screened in one unit or the whole borehole? A pumping test measuring water levels in multiple wells? Via back calculation of water seepage into a long trench or excavation?

Scale effects can be significant.

An SPT in a machine augere hole is a pretty poor-quality undisturbed sample...and what if you have joints at a multi-meter spacing that are significant relative to the work you are doing but you drilled a borehole in a way to miss a bunch of them? I.e. in between joints and the machine borehole smeared it all up and the SPT missed the joints.

A succinct presentation by Martin. Permeability is the most variable soil parameter and still engineers need to assume a value/range based on often commercial k test results in order to proceed with their projects. The tips on this site is really helpful to tackle with the challenges. Thanks.