long-term pumping test Part 2 example

This video describes the analysis of a 38-hr-long well test at the Clemson Gardens wellfield. A conceptual model is outlined and then several analyses from Aqtesolv are described.

[@SharkMerrill]

Thanks Larry for all the videos! I'm having a really hard time figuring out how to get hydraulics conductivity from aqtesolv. Specifically I don't know what to use for B in a lot of cases especially in unconsolidated sediment with variable thickness observed in observation wells. It seems like you usually use aquifer thickness for B. I have a lot of colleagues who say that using the open hole/screen interval is correct. If you have are calculating K from an observation well do you use the aquifer thickness at that location or use the thickness logged in the pumping well? Do you have any tutorials on it or suggested readings on that topic? Thank you so much!

[@lcmurd1]

Mark, Most of the solutions in Aqtesolv will give you transmissivity and you need to come up with a saturated thickness to convert to hydraulic conductivity. The best data would be core or a well log along with the water table. The open hole/screen interval is definitely not the same as saturated thickness, although sometimes that is all the info that is available. Most solutions in Aqtesolv assume the aquifer properties are uniform, but this is never really the case. A good approach is to estimate a range of values for thickness and calculate Ks for the range. Then you can select a value in the middle of the range (geometric mean is often used for K) as your estimate of the uniform value, but you use the range to characterize the uncertainty of that estimate. This would give an uncertainty in K that resulted from uncertainty in the conceptual model (your assumption of uniform thickness). So, you would do your best to come up with a representative value, but you would include the uncertainty in that value. If you have many measurements, you might be able to use more formal statistical methods to characterize a central value and uncertainty. Larry

[@SharkMerrill]

@@lcmurd1 Thank you for the response, I really appreciate the detailed explanation! Your explanation makes a lot of sense and I will definitely use this a lot in the future.

[@ItMaker5000XL]

Larry I love your videos. I work as a hydrogeologist and these videos are such good refreshers on topics I haven't seen in a while, sometimes since school. Thanks so much for putting them out into the world!! Is it possible that the fractured nature of the rock causes the boundary to seem closer, and that's why AQTESOLV seems to only fit well when the boundary is assumed to be closer? If you had a reasonably continuous fracture system, it might evacuate the fractures quite quickly and then hit the stream boundary. That was my first thought on the flatlining of the drawdown. You should go out and pump it for a week to get to the bottom of this!!

[@lcmurd1]

Hi ItMaker, Good point, it is certainly possible that there is a fracture with a high hydraulic diffusivity between the well and the stream. That would help explain the relative difference between the drawdowns in the two wells that are equals distances from the pumping well but different distances from the stream. The analytical solutions are quite idealized and they only allow simple interpretations, but it could be possible to evaluate what you describe using a numerical model that includes discrete fractures. I thought the saprolite was playing an important role in the flattening of the slope, but if that was the explanation I would expect to see drawdown in the saprolite. The data from the saprolite well are peculiar, with some drawdown occurring early and then recovering. These data were measured using a water level tape by different students, so maybe some measurement inconsistencies contributed to the peculiar signal. Agree that a week-long test using pressure transducers would help resolve some questions. Larry