Testing of the IGARF1 v4 spreadsheet tool for assessing the impacts of groundwater abstraction on river flows

This report describes the testing of the IGARF1 v4 spreadsheet tool, developed by the Environment Agency, for the assessment of the impacts of groundwater abstraction on river flows. This spreadsheet, constructed in Microsoft Excel, calculates the effect that a constant or periodic groundwater abstraction has on a groundwater system containing one or two straight-line rivers. The testing involves the simulation of a number of different aquifer configurations to examine if and when the tool has difficulty in calculating a solution. In addition to examining the robustness of the spreadsheet, a number of comparisons are made with a numerical model. There is close agreement between the results produced by the numerical model and the IGARF tool in these comparisons. Whilst the testing has shown that the spreadsheet is a Powerful and easy to use modelling application, that generally produces accurate results rapidly, a number of problems have been encountered. One problem appears to relate to the execution of the dynamic link library (dll) and it is suspected that an error in the dll can cause the tool to crash. This has occurred approximately 6 times during this investigation and appears to be due to memory referencing or allocation errors. However, it is not absolutely certain that the error lies within the dll and it could be a problem with the code contained in an Excel macro. This problem should be resolved before the tool is released to organisations external to the Environment Agency. Four models have been constructed using both the IGARF1 v4 spreadsheet and a numerical model and a comparison has been made between the results of each. The results are in close agreement, except in one example where, as expected, the river becomes perched due to abstraction. The use and testing of the spreadsheet has illustrated that in most cases it returns accurate results. In general it is straightforward to obtain the correct results, however, in approximately half of the model runs, the Scale Factor and Power parameters, on the rdf functions sheet, had to be adjusted to resolve the warning messages that are presented after the solution is calculated. In most cases the warning messages can be resolved quickly, generally by adjusting the Scale Factor parameter only. However, for some models it was not possible to resolve the warning messages by adjusting either or both of the Scale Factor and Power parameters. This problem occurred in examples with realistic parameters and did not just occur in aquifers with unrealistic data sets as expected.