Delay and Loss of Production Water along the Orkla River

It has been found that a simple fixed delay of the water along a river does not provide accurate enough river flow rates for scheduling of the river power plants. One problem is that production water seems to disappear along the river stretch. Another problem is that the delay time depends on the state of the river and is longer in dry than wet conditions.The first problem has been that only 50-80% of the production water released into the river actually arrives at the downstream power plant (after some delay). The mechanism responsible is identified as loss of river water to the sandy valley-bottom aquifer. While most of the production water is delayed with a few hours river delay, the lost fraction is delayed by days and weeks in the aquifer. A coupled tank model is proposed to describe this mechanism. This model has been able to explain the water flow variation due to upstream production variation.The problem with river state dependence of delay has been approached with the method of characteristics, which provides a fast calculation of water wave movement along the river. The river flow rate is specified as a non-linear function of river water content. A two parameter power law is assumed to be valid for the entire river stretch. With this method, it is possible to reproduce the state dependence observed in the river. The studied stretch has a typical delay of 12 hours, but normal variation in the river state can change the delay by several hours. The improved delay calculation probably contributes to reduced production imbalances in the downstream power plant.