GSA Annual Meeting in Indianapolis, Indiana, USA - 2018

Paper No. 280-1
Presentation Time: 1:35 PM

THE EFFECTS OF CHANGES OF LONG-TERM PUMPING ON WETLANDS AND A RIVER IN MICHIGAN


GUST, Spencer Paul and HAMPTON, Duane, Department of Geological and Environmental Sciences, Western Michigan Universiy, 1903 West Michigan Ave, Kalamazoo, MI 49008-5241

The purpose of this study is to forecast the effects that changes in the pumping rates from a major manufacturing plant will have on a river and the intervening land. In a simulation model, we focus on water levels (and flow rates) in the river and areas that were previously wetlands but are now developed.

The plant was built and started up over 60 years ago. Processes in the plant were cooled using groundwater pumped from nearby wells. Over time, pumping increased to over 1.2 m3/s (28 million gallons/day) for both the cooling processes and to maintain hydraulic control of contaminant plumes on site. We are working under the premise that pumping likely diminished the flow of a river a few miles away and lowered water levels between the plant and the river, an area that had been wetlands and farms. Pumping has diminished over the last 20 years by about half due to plume attenuation and process changes.

What effects would be evident in the hydrologic system if the pumping rates and locations changed? If the water pumpage decreased, how would that manifest itself? Presumably, eventually flow in the river would increase and water levels between the plant and the river would rise. This is of interest because the former wetlands area is now dotted with houses and businesses.

To help answer these questions, a three-layer transient MODFLOW model of the glacial geology in a 6 mi2 study area was created. The manufacturing plant has given us access to pumping and hydrogeologic data, including recharge and water pumped off site. Geologic data was collected from a multitude of bore holes and well drilling logs. Other hydrogeologic data provided by the manufacturing plant includes pump tests and pumping records. The model was calibrated using data spanning from the late 1980’s to 2017. Older data (pre-1980’s) was used in building the model, but it was not detailed enough to use in the calibration.

We are studying 5 plausible scenarios in which pumping rates and locations would change over time, all the way to total shutoff (simulated plant shutdown). The model then estimates their effects on water levels and flow in the river. This is an ongoing study.