An evaluation of three methods for assessing long-term well yield

Safe yield and sustainable yield are highly relevant topics in the fields of hydrogeology and water management. Most published information on safe yield has focused on aquifer systems, while relatively little attention has been paid to the safe yield of individual wells, termed here the long-term well capacity (LTWC). The little-studied long-term well capacity is typically viewed as a fixed value, although it might change over the life of any production well.;To test the hypothesis that methods to estimate LTWC may be more broadly applied, three LTWC estimation methods, all based on the theory of radial flow to wells, are tested in a comparative analysis. The methods are tested along with traditional analytical approaches for interpreting pumping tests to determine both hydraulic properties and LTWC. To gain insights into how individual well capacity may be related to the larger context of groundwater management, pumping test data from ten wells completed in a variety of hydrogeologic settings with yields ranging from 2 gpm to 3,500 gpm are assessed using the three LTWC methods, which allow the analyst to estimate either a 100-day (Q 100d) or a 20-year (Q20) LTWC.;The subject wells, completed in confined, unconfined and fractured bedrock aquifers located in Kansas, Oregon and British Columbia Canada, are analyzed using pressure derivative analysis to determine radial flow conditions, interpreted using conventional pumping test analysis to estimate aquifer properties, and then further evaluated using the three methods. From each set of results ratios of projected long-term specific capacity to actual measured short-term specific capacity are derived. A sensitivity analysis performed on the results involved systematically varying pumping time and a default safety factor applied in each LTWC equation. For two wells completed in the Ogallala aquifer in Kansas, the effects of aquifer depletion on LTWC are investigated, and the effects of well interference on two confined aquifer wells in Oregon are assessed.;The first of the three methods tested (Farvolden) was found to produce the least reliable results due to the assumption that drawdown during the first minute of pumping is negligible. This may lead to either an over- or under-estimation of LTWC. The Moell and CPCN methods were found to produce consistent results. Specific capacity ratios derived from each calculated data set helped validate results, and showed that four Farvolden calculations were questionable, projecting a long-term specific capacity that surpassed the measured or observed value during the test. The specific capacity ratios also facilitate qualitative assessment of sensitivity to changes in pumping time or safety factor. For most wells, the results showed similar sensitivity to order-of-magnitude changes in pumping time or 0.1 adjustment in safety factor, with either adjustment changing the resulting LTWC value by approximately 15%, although the LTWC values for the Moell method varied over a wider range when pumping time was varied by an order of magnitude, with the variation depending on whether the well had a low, moderate or high specific capacity ratio.;Effects of aquifer depletion in the High Plains/Ogallala aquifer significantly reduce the LTWC of production wells. This is due to reduced available drawdown (saturated thickness), and decreased aquifer transmissivity from progressive dewatering of permeable aquifer materials. Observed declines in two areas of Groundwater Management District 3 over the past 30 years resulted in an estimated 25 to 30% reduction in transmissivity, and a corresponding decrease in LTWC ranging from 40 to 75% in the two wells studied. The assessment suggests that by 2030 one of the wells studied could have a LTWC value approaching zero.;It is relatively easy to estimate LTWC values for virtually any well subjected to a pumping test without altering the standard procedure for the pumping test. Such an approach coupled with a calculation of the applicable specific capacity ratio supplements the understanding that may be derived from traditional pumping test interpretation. Pumping test data from the time of original well construction are applicable to assessments of current or future LTWC using information on groundwater levels, well specific capacity, pump performance, and other pertinent data.;The evaluation of individual well long-term capacities is recommended when a water source is of particularly high value and there is a need to understand the long-term operational characteristics. Such assessment may be of value in supporting water balance studies in the groundwater systems and the "safe" or "sustainable" yield of aquifers. A Q20 analysis is probably better-suited to municipal wells or other sources used year-round on a perennial basis, while a Q100d analysis may be better suited to wells used intermittently or seasonally for irrigation and other purposes.