Finding Water for Life...

What I hope to do for a living - is work on water resource development projects. I guess I consider myself to be a water resource specialist of sorts (but one with little experience!). Below is a summary of my research work. As I had hoped, after my MBA, I entered the world of environmental consulting and have been loving my work ever since.

A Probabilistic Approach to Ground Water Resource Development in the Carolina Slate Belt.

Selecting sites to drill wells for moderate to high yields in fractured-rock terranes is an inexact science. Yields of individual wells cannot be predicted prior to drilling. As a result, ground water is not usually considered for large supplies in regions underlain by fractured-rock aquifers. However, this problem is becoming acute in the Carolina Slate Belt of Central North Carolina (Guilford, Alamance and Orange Counties) as industrial development, urbanization and population growth continue. The demand for water supplies continue to rise as a result, requiring the need to consider all available resources. Further development of surface supplies is becoming less desirable and more costly. As such, development of the ground water resources in the region is an alternative (Daniel, 1989). A key to this is the ability to identify locations that can support sustained extraction at the necessary rates. In fractured rock terranes, multiple well systems will be necessary to meet large demands. The probability of a given number of wells meeting a specific demand needs to be determined for planning and cost analysis. The associated errors with any prediction also need to be known to establish a range for the number of wells that may be have to be drilled. A statistical approach to quantify the likelihood of obtaining a desired yield, and the number of wells needed to produce a large total yield will be a valuable too for agencies, companies or individuals planning to use ground water from fractured aquifers. The statistical approach will be formulated with as high a degree of translation as possible to similar geomorphic/geologic settings.

The objectives of this project were two-fold: first, to develop a statistical procedure to predict expected well yields and errors associated with any prediction from multiple well systems; second, to apply this procedure to a set of well data with the goal of predicting yields that might be expected from new wells drilled in the same fractured-rock terrane. The analysis was refined by sub setting the well data based on geomorphic criteria.

The study area transects the Carolina Slate belt in the Piedmont Province of North Carolina. It covers most of Orange, Alamance and Guilford Counties and is bounded by the Durham Triassic basin on the East and by the Charlotte Belt on the West. The data to used are well inventory data for each county in the study area. Rather substantial databases for Guilford and Alamance Counties already exist and are maintained in the USGS's Ground Water Site Inventory (GWSI) database. Orange County, however, lacked a sufficient well inventory. To overcome this lack of data, inventory of new well data including field-checking of locations, geology and well construction was conducted. The well inventory of Orange County comprised the fieldwork for this study. The numerical data was then statistically analyzed, including sampling with and without replacement in subsetted and non-subsetted groups based on associated drainage patterns, and varying numbers of wells.

The maps to the left indicate the location of the study area.

Current Research Status

1. Well inventory and logging COMPLETE. Wells from previous workers and current work obtained and tabulated. Dataset includes well logs from Bane (1969), Sharpless (1985), McKelvey (1994), Wallace (1996), and current logs from the Orange County Planning Department (Jones and Ali, 1997).
   

2. Digitizing of well data COMPLETE. Wells not in digital form digitized and plotted onto 1:24,000 scale 7.5"; topographic quadrangle maps.
   

3. Geomorphic analysis COMPLETE. Approximately 10 geomorphic characteristics obtained from the plots of the wells on the topographic quadrangles are being entered into a new database for approximately 800 wells lacking such information. Geomorphic data include well elevation, nearest stream elevation, geologic setting, topographic setting, texture of underlying rocks, orthogonal ray from well to nearest stream, ray of nearest stream, distance from well to nearest stream and hydrological unit in which well is situated.
   

4. Development of SAS statistical analytical procedure COMPLETE. SAS code needed for a custom statistical analytical procedure developed, including the sorting and merging of several datasets,  random number generation and sampling without replacement with filters to minimize duplicate observations.
   

5. Statistical Analysis COMPLETE. SAS procedures applied to dataset. Completed analyses include UNIVARIATE, MEANS, ANOVA and other descriptive statistics. Procedure to predict expected well yields and errors associated with any prediction from multiple well systems that might be expected from new wells drilled in the same fractured-rock terrane applied to entire dataset and subsets.  Probability distribution of well yields from one or multiple boreholes for entire dataset and subsets of Topographic setting and Drainage Control described and graphically represented.
   

6. Written Thesis COMPLETE, DEFENDED, SUBMITTED AND ACCEPTED!.
   

7. Presentation to Geological Community COMPLETE.
   

As you may have read in my "about me" page, I presented my work at the 1998 Annual Geological Society of America Meeting in Toronto, Canada. An off-shoot of this work was completed by Dr. Charles C. Daniel, III (my Advisor at UNC) looking only at the commercial wells. The results were spectacular, and confirmed the findings of my thesis work. These results were presented at the 1999 Annual Southeast Regional GSA Meeting in Athens, GA.

I am planning to expand this work to a broader geological region and hopefully complete my PhD as a result. However, this will take time, so check back in a few years!