DMOS6 Chiller Plant Optimization

Photo of José Miguel Peña, Daniel Richards (Team Leader), Brian Smith Students: José Miguel Peña, Daniel Richards (Team Leader), Brian Smith

Sponsor: Texas Instruments

Date: Fall 2007

Reliability is paramount; it is crucial that the chiller plant remain on and properly running at all times. Texas Instruments can suffer up to $5 million in daily losses if the DMOS6 chiller plant fails. An economic analysis must be performed on any recommended design changes to ensure a savings payoff within two years. Design changes may include but are not limited to the addition of cooling towers, bypasses, and new pumps. Beyond the economic constraints, there are also several engineering constraints. First, the pressure drop across the three main feed/return lines must be kept at 20 psi. Also, the flow rate through the chillers must remain above 1200 gpm to prevent the refrigerant from freezing. Furthermore, the temperature of the water leaving the chillers must remain at a constant 42 degrees Fahrenheit as specified by TI.

Our task was to provide cost-effective and implementable chiller plant operation strategies and reconfiguration concepts to TI based on investigations made using computer models and by examining the effects of operation strategies on the actual DMOS6 chiller plant.

The intent of this project was to provide Texas Instruments with recommendations to increase the efficiency of the DMOS6 chiller plant. We accomplished this by investigating optimized operation strategies, component reconfiguration, and research of the available literature. We recommend that TI make three main operational changes regarding the chiller plant. First, the chillers should be run as close to their optimal %RLA as possible. Deviating from the optimal state both increase the operating cost of running the chillers and decreases the efficiency of the chillers. Second, the chillers should be turned on one at a time as opposed to a pair. The chillers consume the most power in the chiller plant and should be run with excess capacity. Third, the condenser water should be run at a high flow rate. Based on an analysis of the plant under the most and least efficient operation of the plant; a thermodynamic first law analysis; comparison of the test run to similar data last year; and the simple heat rejection study, it is much more efficient to run the plant at a high condenser water flow rate per chiller. We further recommend that TI continue investigation into increasing chiller plant efficiency based on the preliminary analysis of the secondary pumps. First-order analysis shows that it may be possible to run the chiller plant without the 5 primary pumps. Further investigation is needed beyond the scope of this project to determine if running in this configuration will increase plant efficiency and whether the chiller water flow can be controlled using only the secondary pumps.

Images related to the project:

Photo related to DMOS6 Chiller Plant Optimization project
Joomla SEF URLs by Artio