Low Delta-T Syndrome Diagnosis and Correction for Chilled Water Plants

This thesis presents a study of the low delta-T syndrome in the DFW Airport. The chilled water system was modeled separately to understand its behavior and identify the possible causes for the low delta-T. The aim of this study is to propose energy saving measures that would improve the plant?s overall efficiency.

Data for the chillers and the chilled water storage tank was analyzed to determine the DFW Airport?s chilled water plant operation. The chiller simulation, following Gordon and Ng?s model, provided the starting point to the optimization process. The chiller model?s input variables were chilled water leaving temperature, condenser water entering temperature and water flow.

A detailed study of the TES tank was made to ensure that the proposed changes would enable the TES tank to provide cooling for the airport, particularly during the summer months from June through September from 3 pm to 6 pm when the chillers are turned off due to the rate structure.

The chilled water leaving temperatures were set to provide enough chilled water to meet the cooling loads and provide the necessary chilled water to fill the TES tank. The other variables, namely condenser water leaving temperature and water flow, do not affect the rate at which the tank is filled. Thus, the values obtained through the optimization process are maintained.

The proposed reset schedule for the DFW Airport is the following: constant condenser water flow of 100,000 ft^3/hr (~12,500 gpm), an approach temperature of 5?F whenever the outside air wet bulb temperature is above 55?F, otherwise maintain a 60?F condenser supply temperature. The reset schedule proposed for the chilled water supply temperature is 37?F from March through September and 42?F from October through December, January and February.

The chiller simulation and the secondary pumps simulation were used to determine the power required to run the system under the proposed reset schedules. Initially, the condenser water flow rate was set to current operation because it was not possible to predict the additional power required by the condenser pumps. A condenser pumping power model was not built due to lack of data. The total energy savings predicted were 1,589,000 kWh, which represents 3.8% savings. Finally, the simulation was run using the proposed condenser water flow rate of 100,000 ft^3/hr (~12,465 gpm) per working chiller, year round. The predicted energy savings were 3,237,000 kWh, which represents 7.7% savings. However, the savings will be less than 7.7% when the additional power required by the condenser pumps is considered.