Date of Award


Document Type

Open Access Dissertation

Degree Name

Doctor of Philosophy (PhD)


Environmental Sciences/Environmental, Earth & Ocean Sciences

First Advisor

Anamarija Frankic

Second Advisor

Ellen Douglas

Third Advisor

David Terkla


Large thermoelectric facilities are issued permits to discharge high volume, high temperature effluents as part of the National Pollutant Discharge Elimination System (NPDES). Once-through cooled power plants are especially dependent on large quantities of cool water to operate. When ambient temperatures are high or streamflow is very low, power plant managers must reduce (i.e., "dial back") energy generation in order to avoid violating their NPDES permit limitations. Sudden dial-back can have human health impacts when electricity is no longer available to provide cooling or other vital services. A superior system of electricity and environmental management would reduce the probability of future violations and/or dial-back by explicitly recognizing the facilities for which those events are highly likely. An original statistical model is presented and used to answer the following research questions: 1) Do electricity demand and natural environmental conditions influence withdrawal rates and effluent temperatures at once-through thermoelectric facilities? 2) Is it possible to estimate past withdrawal rates and effluent temperatures where reported observations are unavailable? 3) In the future, how often will power plant managers face the decision to dial-back generation or violate their plant's discharge permit? 5) What can be done to avoid such decisions and the resulting negative impacts? Two facilities in Massachusetts were chosen as representative case studies. Using public records, several decades of daily and monthly observations of environmental variables (e.g. ambient air temperature, streamflow) and monthly energy generation were tested against monthly observations of facility water withdrawal rates and maximum discharge temperatures using a multiple linear regression (MLR) approach. The MLR model successfully estimated monthly maximum discharge temperatures for both facilities using monthly average of daily high air temperatures and monthly net electricity generation. The model was used to identify months in the past when violations or dial-back are likely to have occurred, as well as months in the future when each plant is expected to dial-back or violate its permit as ambient air temperatures continue to rise. Solutions are presented that reduce the number of predicted violations, meet consumer electricity demand to the greatest extent possible, and reduce the chances of sudden dial-back.